Generics (UK) Ltd v Daiichi Pharmaceutical Co Ltd & Anor

In this action the claimant (“GUK”) seeks a declaration of invalidity of (or rectification of the register in respect of) supplementary protection certificate no. SPC/GB97/085 (“the SPC”) in the name of the second defendant and a declaration that claims 1-2 and 5-7 of European Patent (UK) No. 0206283 (“the Patent”) in the name of the first defendant were invalid. I refer to the defendants collectively in this judgment as “Daiichi”.

The Patent relates to the (-) enantiomer of a racemic compound called ofloxacin. Ofloxacin is a member of the quinolone class of anti-microbial agents. The (-) enantiomer of ofloxacin is called levofloxacin and its structure is shown below, with an asterisk marking the chiral centre:

It was agreed at trial that the claims of the Patent upon which I need to focus are:

The Patent claims priority from three Japanese filings dated 20 June 1985, 11 October 1985 and 28 January 1986, each of which describes a process for making levofloxacin. They are referred to in the Patent as processes A, B and C respectively. The application for the Patent was filed on 20 June 1986 and it proceeded to grant on 27 January 1993. On 6 June 1997, UK marketing authorisations were granted in respect of levofloxacin. On 23 October 1997, Daiichi lodged the application for the SPC, identifying the Patent as the basic patent, levofloxacin as the product and the UK marketing authorisations in respect of levofloxacin as the first authorisations to place the product on the market. The SPC was duly granted on 13 July 1998. The Patent expired on 20 June 2006, but the SPC is now in force and, subject to this action, will expire on 19 June 2011.

The validity of the SPC is challenged in two distinct ways. First, it is said that the SPC is invalid pursuant to Article 15 of Council Regulation (EEC) No 1768/92 (“the SPC Regulation”) because grounds exist which would have justified revocation of the Patent, or its limitation to the extent that levofloxacin would no longer have been protected by the claims. Second, there is a free standing attack on the SPC (or its duration) based upon earlier marketing authorisations for ofloxacin.

The issues in the action which occupied most of the time at trial were those concerning the validity of the Patent. They are:

Finally I would note that GUK has sought to establish lack of novelty and obviousness over all the above prior art, save for Gerster IP, by performing experiments purporting to show obvious ways in which the skilled person could have resolved ofloxacin into its enantiomers in 1985. These experiments were the subject of a good deal of evidence and criticism, as I shall explain.

Two experts gave evidence on behalf of GUK, Dr Peter Spargo and Dr Roger Newton.

Dr Spargo gave the primary evidence on the issue of obviousness over the various citations relied upon by GUK. He obtained a First Class Honours degree in Natural Sciences (Chemistry) from Cambridge University in 1983 and was awarded a PhD in Synthetic Organic Chemistry in 1986, also from Cambridge University where he held a non-stipendiary research fellowship. After two years as a NATO Postdoctoral Research Fellow and a Lindemann Trust Fellow at Columbia University, he returned to the UK, joining Pfizer Ltd (Sandwich) in 1988 as a medicinal chemist. Within two years he transferred to Pfizer’s Process (now Chemical) Research and Development Department, where he progressed from Laboratory Team Leader to Section Head, then Manager, Director, and ultimately Head of Department (Group Director). During his 15 years at Pfizer, Dr Spargo identified, developed and scaled up a wide range of manufacturing processes for new drug candidates across a number of different therapeutic areas.

Dr Spargo gave his evidence in a forthright, clear and balanced way and he sought to retain his objectivity throughout his extensive cross examination. However it is notable that Dr Spargo was studying for his PhD at the 1985 priority date, did not profess to be an expert in medicinal chemistry and, as he frankly accepted, had no experience of quinolone chemistry before his involvement in these proceedings. It is also relevant that he formed his views of the Patent and the prior art before having read any background materials that would have permitted him to gain an understanding of the common general knowledge in the quinolone field. Nor did he claim to be an expert in HPLC chromatography, an issue of some importance in relation to the experiments. All these are matters which I must take into account in considering the weight to be given to his evidence when assessing obviousness.

Dr Spargo was also subjected to more severe criticism for adopting what was said to be an unreasonably low standard in assessing what amounted to a reasonable expectation of success, displaying an unreasonable determination to want to obtain the enantiomers of ofloxacin and, perhaps most seriously, for revealing a desire to argue GUK’s case. I reject all these criticisms. I did not regard any evidence given by Dr Spargo as unreasonable. Nor do I believe he was attempting to argue GUK’s case. He expressed his opinions at all times in a measured and sensible way. As I have said, I must consider the weight to be attached to those opinions in addressing the obviousness allegation. But I have no doubt they were all expressed with the intention of assisting the court.

Finally it was said Dr Spargo played an unsatisfactory role in connection with the experiments, that his report was not complete and his advice was deployed in a partial manner. This arose from disclosure given in the course of the trial of attendance notes and other correspondence containing details of the instructions given to and advice received from Dr Spargo as to the reagents and techniques he would suggest for the resolution of ofloxacin. I deal with the experiments later in this judgment and for the moment limit myself to the personal criticism directed at him. I found Dr Spargo to be completely candid in the witness box and I do not believe for one moment he attempted to deceive or mislead the court. In so far as his report gave a less than full account of all his instructions and advice, I am sure that was because he believed he had disclosed all that was relevant. Once again, I reject this criticism.

Dr Newton obtained his BSc in Chemistry and Biology from the University of London and was awarded his PhD in 1968. From 1991 to 1996 he was employed by Glaxo, becoming, from 1983-1988, Director of Chemistry and, from 1989-1995, Director of the Chemical Research Division. During that time, Glaxo discovered and launched a number of very successful pharmaceuticals including Imigran, Zofran and Serevent. Whilst at Glaxo, Dr Newton acted as Deputy Director of the Anti-infective Research Group which was involved in antibiotic research but with a focus towards the cephalosporins rather than quinolone based antibiotics.

A good deal of the evidence given by Dr Newton related to the basic concepts of stereochemistry, about which there was no dispute. However, he also addressed the attitude of the pharmaceutical industry in the mid 1980s to racemates. Here he suffered from the same difficulty as Dr Spargo in that he had no experience of the quinolone field and, again, this is a matter I must take into consideration when considering obviousness. It was also suggested that various detailed points of Dr Newton’s evidence were unsatisfactory in that, for example, he had no knowledge of specific aspects of ligand-receptor binding and had overlooked a feature of the configuration of one of the quinolone molecules the subject of evidence. In my judgment matters of this kind form no basis for making general criticisms of an expert and I reject the attack. Dr Newton gave his evidence clearly and fairly.

Daiichi relied on the evidence of three experts, Professor Mark Wentland, Professor Stephen Davies and Professor George Zhanel. In addition Daiichi called a witness of fact, Dr Judith Weissinger, who was, at the relevant time, at the US Food and Drug Administration (the “FDA”).

Professor Wentland joined the medicinal chemistry department at the Sterling-Winthrop Research Institute in Rensselaer, New York in 1970. In the period 1981 to 1990, he became heavily involved in quinolone-related research as a synthetic medicinal chemist. During much of that time, he was co-project leader for the quinolone discovery team, responsible for running the chemistry parts of the program. His co-project leader was a microbiologist. Professor Wentland explained that he made significant contributions to the discovery of nine development compounds, six of which entered clinical trials. One of these was amifloxacin, a quinolone antibacterial that entered Phase II clinical trials. In 1988, Sterling-Winthrop was bought by Eastman-Kodak and the antimicrobial research stopped. Professor Wentland continued to work on quinolones at Sterling-Winthrop after this time, but for the purpose of finding an anti-cancer drug rather than an antimicrobial. During the period 1971 to 1994, he held the concurrent position of Adjunct Professor of Chemistry at Rensselaer Polytechnic Institute where he taught many graduate-level courses in organic and medicinal chemistry.

Professor Wentland was accused by GUK of giving evidence which was, on occasion, inexplicable and of lacking objectivity and independence under cross examination. I am unable to accept these criticisms. I believe Professor Wentland gave his evidence in a careful and measured way. Importantly, he was the only expert to give evidence before me who was actually involved in quinolone research at the priority date. I recognise that Professor Wentland’s views of the common general knowledge and the attitude of the skilled person were founded very largely on his experience at Sterling-Winthrop and I must therefore be wary of drawing general conclusions. Nevertheless, I found his evidence of great assistance and his opinions carry considerable weight.

Professor Davies is the Chairman of Chemistry and the Waynflete Professor of Chemistry at the University of Oxford. He obtained his B.A. in 1973 and was awarded a PhD in 1975, both in Chemistry from the University of Oxford. He received a D.Sc. degree in Chemistry from the University of Paris in 1980. His PhD was entitled "Studies on Epoxides" and his D.Sc. was entitled "Contribution a La Chimie des Epoxydes" both of which described the stereoselective synthesis and reactions of epoxides. He has published widely and has won a number of prestigious awards and prizes. His research interests since 1973 have been generally in the area of the stereochemical aspects of organic and organometallic chemistry, and the preparation of enantiomerically pure materials. His particular interests are asymmetric and stereoselective syntheses of homochiral compounds. He founded Oxford Asymmetry in 1991, with the aim of providing pharmaceutical companies with homochiral compounds of interest on any desired scale.

The principal criticism of Professor Davies was that he is not a medicinal chemist and has never worked in industry. As a result he pitched the knowledge and experience of the notional skilled medicinal chemist at too low a level. Moreover, it was said, he adopted an unremittingly negative approach to the attitudes and abilities of the skilled chemist and to the teaching of the cited prior art. I think these attacks are unduly severe. Professor Davies is an eminent organic chemist with immense experience of stereochemistry and resolution techniques. He was well placed to assist me in relation to the chemistry aspects of the case, as to the predictions that could or could not be made from structures and as to the likely success of different resolution techniques. Moreover, he had contacts with industry through a number of consultancies with leading pharmaceutical companies, including Fisons, ICI Pharmaceuticals and Hoechst, who have consulted him on a range of problems. In my judgment Professor Davies’s evidence was clear and cogent. But I recognise that he neither worked in industry nor had direct experience of the quinolone field and these are matters which I have had well in mind when assessing the weight to be attached to his opinions.

Professor Zhanel is a Professor in the Department of Medical Microbiology, Faculty of Medicine, University of Manitoba and is currently Chair of the antibiotic resistant infections research group. He is a Doctor of Pharmacy and gained his PhD in Medical Microbiology at the University of Manitoba in 1994. The subject of his doctoral thesis was "Cellular and Molecular Evaluation of Fluoroquinolone Resistance in Pseudomonas aeruginosa". Since about 1985 the focus of his work has been anti-infectives, including quinolones. He has considerable practical experience of the treatment of patients with infectious diseases.

As GUK recognised, Professor Zhanel is clearly a huge enthusiast for levofloxacin and a great believer in its superiority over ofloxacin. However, it was submitted, with some justification, that aspects of his evidence were somewhat exaggerated. For example, he said that in 1985 it was clear that ofloxacin did not represent an agent to be used for the treatment of community acquired respiratory tract infections. However, it was the subject of clinical trials for pneumonia and chronic bronchitis from 1985 and these provided the summary basis of approval of the drug by the FDA for the treatment of adults with these infections. He also said that the whole infectious diseases community was furious with the inclusion of ofloxacin as a fluoroquinolone with enhanced activity against S. pneumoniae in the 1998 guidelines of the Infectious Diseases Society of America (“the IDSA”). Yet this was a document subjected to external review by peer reviewers and had been approved by the IDSA’s Practice Guidelines Committee and the IDSA Council. It was therefore submitted I should approach Dr Zhanel’s evidence with great caution. I accept I must approach Dr Zhanel’s evidence with care. However I should also say that I found him to be an extremely knowledgeable witness and I think his tendency to exaggeration was the result of over enthusiasm rather than any intention to mislead. Overall, I found his evidence helpful.

Dr Weissinger gave evidence as to the approach adopted by the FDA to the approval of racemic drugs in the mid 1980s. In 1984, she joined the FDA as a reviewer in the Center for Veterinary Medicine in the Office of New Animal Drug evaluation. In 1988 she was re-assigned to the Office of Biologics Research and Review and then, in early 1988, to the Office of Drug Evaluation within the Center for Drug Evaluation and Research (“CDER”). These offices are responsible for the approval of, respectively, biological products and drugs for human use. One of her roles was to serve on the CDER Stereoisomer Policy Committee with responsibility for developing an FDA policy for the approval of stereoisomeric drugs. No criticism was made of Dr Weissinger as a witness, and rightly so.

The specification begins with an acknowledgment that ofloxacin is known to be an excellent synthetic antimicrobial and is disclosed in the 005 Patent. It continues by noting that ofloxacin has an asymmetric carbon atom at the 3 position and is obtained as a racemate by known processes. It then says that the inventors have obtained the enantiomers and found levofloxacin (the S(-) enantiomer) possesses an antimicrobial activity of about twice that of the racemate and an acute toxicity weaker than the racemate, as determined in mice by intravenous injection. On the other hand, the R(+) enantiomer exhibits an antimicrobial activity of only about 1/10 to 1/100 times that of the racemate, with approximately the same toxicity. It concludes that levofloxacin has been found to have very desirable properties, that is to say, increased antimicrobial activity and reduced toxicity, and is expected to be a very useful pharmaceutical agent as compared to the racemate. It continues that both the enantiomers have been found to have markedly increased solubility as compared to the racemate and other free compounds of this type, and can be used in injectable preparations.

The objects of the invention are then described as the provision of an optically active isomer of ofloxacin and its analogues and to provide novel processes for preparing these compounds.

The specification continues with a description of the three processes A, B and C for making levofloxacin and its analogues. Process A is explained at page 4, line 16 - page 5, line 55 and in examples 1-7. In short, it proceeds via a resolution of a covalent dinitrobenzoate ester intermediate using chiral HPLC.

Process B is described at page 6, line 1 – page 9, line 10 and in examples 8-11. It proceeds via resolution of an acetate ester intermediate using a hydrolytic enzyme, such as a lipase.

Process C is described at page 9, line 11 - page 11, line 53 and in examples 12-16. It involves resolution of the intermediate denoted (X’) using N-tosyl-L-prolyl chloride (referred to in Example 12 as the acid chloride of (S)-N-p-toluenesuphonylproline) - the resolving agent used by Riker in Gerster IP and Gerster II - to form the enantiomer (X).

Enantiomer (X) is then converted to levofloxacin (VI) by the same series of reactions used to convert the racemic form of that intermediate to ofloxacin published in Drugs of the Future and the 005 Patent. It is right to note that the Patent itself acknowledges at page 11 lines 50-52 these are “known reactions” and also states at page 13 lines 37-39: “The conversion of the intermediate (X) into the desired ofloxacin or an analog thereof can be carried out by a well-known process as disclosed, for example, in…European Patent Specification EP-B-47005…”

Of all these processes, C is identified as “particularly preferred”.

The antimicrobial activities of racemic ofloxacin, its two individual enantiomers and the S(-) enantiomer of the N-ethylpiperazine analogue of ofloxacin against nine micro-organisms are described in Table 2 on page 12. Levofloxacin is shown to be almost exactly twice as active as the racemate in the tests performed. On the other hand, the R(+) enantiomer is much less active than the racemate. The acute toxicity (LD₅₀) data of the racemic ofloxacin and its two individual enantiomers are detailed in Table 3. It can be seen that levofloxacin is less toxic than the racemate and the R(+) enantiomer at 200mg/kg. The water solubilities of racemic ofloxacin and its two individual enantiomers are described in Table 4. According to this table, the water solubility of each individual enantiomer is approximately the same, and each is approximately 10 times that of the racemate.

It is to be noted that the first priority document refers only to the two fold increase in activity. It does not refer to toxicity or solubility. It is only in the application as filed that the toxicity and solubility data are included. This forms the basis of the attack on the priority date of the Patent.

No issues of interpretation arise in relation to the claims. Importantly, GUK accepted that ofloxacin (the racemate) does not fall within the scope of claim 2, directed as it is to levofloxacin. Beyond this, the precise scope of claim 2 is not relevant to any issue I have to decide. Claim 5 is a claim to a process for making levofloxacin involving adding N-methylpiperazine to the intermediate denoted V on page 28 of the Patent.

A patent specification is addressed to those likely to have a practical interest the subject matter of the invention, and such persons are those with practical knowledge and experience of the kind of work in which the invention was intended to be used. The skilled addressee comes to a reading of the specification with the common general knowledge in the art and he reads it on the assumption that its purpose is to describe and demarcate an invention. He is unimaginative and has no inventive capacity.

There was agreement between the parties that the skilled addressee of this Patent is a team and there was little between them as to its composition. Dr Spargo said it would be a team in a pharmaceutical company dedicated to the discovery and development of antimicrobial agents and, in particular, quinolones and it would include organic, medicinal, analytical and process chemists, a microbiologist and a pharmacologist. The team would also have access to the input of clinicians specialising in the relevant disease areas and a toxicologist. The microbiologist and pharmacologist would advise the team generally and offer opinions from their own perspectives. The microbiologist would have a particular interest in the pharmacological and microbiological tests necessary to establish efficacy. As Professor Zhanel elaborated, the pharmacologist would advise the team as to how a putative drug might interact with the human and its potential benefits and drawbacks in terms of activity, pharmacokinetics, creation of resistance and toxicity. Nevertheless, as Professor Zhanel also agreed, the medicinal chemist would have primary control over the direction of the project.

The dispute, such as it was, concerned the role, if any, of the process chemist, a matter of some significance since that is Dr Spargo’s primary expertise. Daiichi contended the team would not include a process chemist at all because the Patent is, on its face, directed to a new chemical entity. On the evidence as a whole, I believe a process chemist would be part of a skilled team in the quinolone field. The process chemist would monitor the project and influence its direction, not least through advice as to which quinolone targets could actually be made, a matter confirmed by Professor Wentland in his first report. In this regard it is also of interest that the patent includes process claims and describes methods of synthesis. Nevertheless, I am satisfied that the role of the process chemist would largely be one of support to the medicinal chemist.

It is important to have a clear understanding of the meaning of the common general knowledge. The correct approach was explained by the Court of Appeal in Beloit Technologies Inc v Valmet Paper Machinery Inc [1997] RPC 489 at pages 494-495:

“It has never been easy to differentiate between common general knowledge and that which is known by some. It has become particularly difficult with the modern ability to circulate and retrieve information. Employees of some companies, with the use of libraries and patent departments, will become aware of information soon after it is published in a whole variety of documents; whereas others, without such advantages, may never do so until that information is accepted generally and put into practice. The notional skilled addressee is the ordinary man who may not have the advantages that some employees of large companies may have. The information in a patent specification is addressed to such a man and must contain sufficient details for him to understand and apply the invention. It will only lack an inventive step if it is obvious to such a man.

It follows that evidence that a fact is known or even well-known to a witness does not establish that that fact forms part of the common general knowledge. Neither does it follow that it will form part of the common general knowledge if it is recorded in a document. As stated by the Court of Appeal in General Tire & Rubber Co. v. Firestone Tyre & Rubber Co. Ltd. [1972] R.P.C. 457, at page 482, line 33:

"The two classes of documents which call for consideration in relation to common general knowledge in the instant case were individual patent specifications and widely read publications'. As to the former, it is clear that individual patent specifications and their contents do not normally form part of the relevant common general knowledge, though there may be specifications which are so well known amongst those versed in the art that upon evidence of that state of affairs they form part of such knowledge, and also there may occasionally be particular industries (such as that of colour photography) in which the evidence may show that all specifications form part of the relevant knowledge.

As regards scientific papers generally, it was said by Luxmoore, J. in British Acoustic Films (53 R.P.C. 221 at 250):

"In my judgment it is not sufficient to prove common general knowledge that a particular disclosure is made in an article, or series of articles, in a scientific journal, no matter how wide the circulation of that journal may be, in the absence of any evidence that the disclosure is accepted generally by those who are engaged in the art to which the disclosure relates. A piece of particular knowledge as disclosed in a scientific paper does not become common general knowledge merely because it is widely read, and still less because it is widely circulated. Such a piece of knowledge only becomes general knowledge when it is generally known and accepted without question by the bulk of those who are engaged in the particular art; in other words, when it becomes part of their common stock of knowledge relating to the art."

And a little later, distinguishing between what has been written and what has been used, he said:

"It is certainly difficult to appreciate how the use of something which has in fact never been used in a particular art can ever be held to be common general knowledge in the art."

Those passages have often been quoted, and there has not been cited to us any case in which they have been criticised. We accept them as correctly stating in general the law on this point, though reserving for further consideration whether the words 'accepted without question' may not be putting the position rather high: for the purposes of this case we are disposed, without wishing to put forward any full definition, to substitute the words 'generally regarded as a good basis for further action'.”

Thus the common general knowledge is the common knowledge in the field to which the invention relates. The notional skilled addressee is the ordinary man who may not have the advantages that some employees of large companies may have and information does not form part of the common general knowledge simply because it is known to some persons in the art. It must be generally known and generally regarded as a good basis for further action by the bulk of those engaged in that art before it becomes part of their common stock of knowledge relating to the art, and so part of the common general knowledge. That is not to say the skilled person must have it at the forefront of his mind. As Laddie J explained in Raychem Corporations’ Patents [1988] RPC 31 at 40, it includes all the material which he knows exists and which he would refer to as a matter of course if he cannot remember it and which he generally understands is sufficiently reliable to use as a foundation for further work.

GUK suggested that the common general knowledge goes rather further and in this connection referred me to the observations of Pumfrey J in Novartis v Ivax [2006] EWHC 2506 (Pat) at [27]:

“It is well settled that the common general knowledge is knowledge that must be attributed to the skilled person, without which the latter may be taken not to be skilled. To it must be added any knowledge that every skilled person should be taken to acquire before he embarks on the problem to which the patent provides the solution, as, for example, the relevant properties of cyclosporin if the problem is to formulate cyclosporin.”

And again in Glaxo Group’s Patent [2004] RPC 43 at [35]:

“35.

These provisions do not permit what is sometimes called the mosaicing of individual documents or prior uses said to form part of the state of the art, unless it can be shown that the skilled person, confronted with a particular citation, would turn to some other citation to supplement the information provided by the first. Such cases are not common, and the problem most frequently arises when the patent in suit itself makes assumptions about what its disclosure enables a skilled man to do. To take an example, one might consider a prior use consisting of the sale of a particular pharmaceutical preparation. The defendant says that the nature of the active ingredient can be ascertained by analysis of the preparation. Whether that active ingredient forms part of the state of the art depends upon identifying the experimental techniques available to the skilled person. The use of such techniques may involve reference to books and journals, and the explanation of the results may require reference to particular publications. If it can be shown that this analysis is the kind of analysis that would be normally performed, then it is part of the common general knowledge even if only a fraction of the material can be ascertained without reference to the books. Laddie J gives a further example in paragraph 66 of his judgment in Pfizer:

‘When any piece of prior art is considered for the purposes of an obviousness attack, the question asked is “what would the skilled addressee think and do on the basis of this disclosure?” He will consider the disclosure in the light of the common general knowledge and it may be that in some cases he will also think it obvious to supplement the disclosure by consulting other readily accessible publicly available information. This will be particularly likely where the pleaded prior art encourages him to do so because it expressly cross-refers to other material. However, I do not think it is limited to cases where there is an express cross-reference. For example if a piece of prior art directs the skilled worker to use a member of a class of ingredients for a particular purpose and it would be obvious to him where and how to find details of member of that class, then he will do so and that act of pulling in other information is itself an obvious consequence of the disclosure in the prior art.’”

It seems to me that a subtle but potentially significant point of principle emerges from these passages. I can readily accept that, faced with a disclosure which forms part of the state of the art, it may be obvious for the skilled person to seek to acquire further information before he embarks on the problem to which the patent provides a solution. But that does not make all such information part of the common general knowledge. The distinction is a fine one but it may be important. If information is part of the common general knowledge then it forms part of the stock of knowledge which will inform and guide the skilled person’s approach to the problem from the outset. It may, for example, affect the steps it will be obvious for him to take, including the nature and extent of any literature search.

Much of the background was not in dispute and I take it, with gratitude, from the evidence of Professor Wentland and Professor Zhanel. The term "quinolone" is a generic term used to describe a class of antibacterial agent. It is properly meant to describe a 1-substituted-1,4-dihydro-4-oxo-3-pyridinecarboxylic acid that has additional ring(s) fused to it and may have additional nitrogen atoms in those rings. This structure is identified in the diagram in paragraph [42] below as the quinolone core structure. Quinolones with bicyclic core structures where X can be carbon or nitrogen (see the bicyclic core structure below) are strictly considered by Chemical Abstracts Service ("CAS") to be "quinolines" or 1,8-naphthyridines, respectively, and positions on these rings are numbered as shown. For example, nalidixic acid has two nitrogen atoms in its bicyclic core structure making it a 1,8-naphthyridine derivative.

A third ring may be added to provide compounds such as flumequine and ofloxacin. The core tricyclic ring systems of these two agents have a different CAS numbering system (again, see below). Throughout the proceedings the parties endeavoured to use the bicyclic core structure numbering system whenever possible. However, when referring to appendages on the third ring, they tended to use the precise CAS system. For example, Professor Wentland referred to the methyl attached to the centre of chirality of ofloxacin (shown with a star) as the "3-methyl" group. For simplicity, the parties used the term "quinolone" to describe all the compounds depicted and their derivatives and in this judgment I shall do the same.

It is a feature of the quinolones that they do not mimic a natural ligand. However, it has been known for very many years that they have as their target a bacterial enzyme called DNA gyrase. This enzyme is used by bacteria to control the shape of DNA during replication, in the course of which a particular transient adduct (known as the cleavable-complex) is formed. By 1985 it was known that quinolones act by binding to the cleavable-complex and so prevent it from completing its role in the replication process. However, details of the structure of the cleavable-complex and precisely how quinolones interact with it were not understood.

Nevertheless, by 1985 a large number of quinolones of very different structures had been shown to display antibacterial activity. This led to the adoption of the “induced-fit” model to explain the binding of quinolones to the cleavable-complex. This model suggests that flexibility in the shape or conformation of the molecular target allows different drugs to bind to the same site by inducing a good fit. It is to be contrasted with the “lock and key” model in which the target (the lock) retains its shape irrespective of the structure of the drug molecule (the key).

The process of assessment of a putative antimicrobial drug in June 1985 was described by Professor Zhanel. First, a clinical pharmacologist or microbiologist would look at the minimum inhibitory concentration (“MIC”) for the compound of interest and a particular bacterium. The MIC is the lowest concentration of a compound which inhibits bacterial growth. The lower the MIC, the greater the activity. This MIC data would then be compared to a standard antimicrobial in the same class.

DNA gyrase assays were also used in some companies at this time. These measured the concentration of a compound required to inhibit the activity of the DNA gyrase. If the compound did not display promise in standard MIC or DNA gyrase tests, then, at least in general, no further MIC testing or in vivo (animal efficacy or pharmacokinetic or toxicity experiments) tests would be performed. On the other hand, if the compound looked promising it would undergo further in vitro and in vivo testing.

In 1985, the standard assay to assess antibacterial activity in vivo was known as the mouse protection model. In summary, animals are injected with a bacterial inoculum and deaths are recorded over a period of days. Other groups of animals are provided with different doses of the putative antibiotic a short time after infection and fifty percent protective dose values (PD₅₀) are then calculated. This represents the dose that protects 50% of mice in a test group from an otherwise lethal infection. Likewise, toxicity of a drug is assessed by the dose needed to kill 50% of animals in the test group (LD₅₀).

An important further step for a promising candidate was to consider the peak serum and urinary concentrations of the compound obtained in humans after standard dosing. These would then be compared to MICs of common infectious pathogens. For systemic infectious diseases, the objective was to find a compound with a peak serum concentration of at least four times the MIC. It was understood that peak serum levels decline rapidly and a level of at least four times the MIC was thought necessary to eradicate all the bacteria and prevent the development of resistance.

For the treatment of urinary tract infections, bacteriological and clinical cure is poorly related to achievable antimicrobial concentrations in the blood. Rather, bacterial eradication from the urine is correlated with antimicrobial concentrations achieved in the urinary tract and it was believed that efficacy required a minimum of 30% of the active compound to be excreted in the urine.

Other significant properties of an antibiotic were also explored during the trial, and again were the subject of an explanation by Professor Zhanel, which I gratefully adopt. Pharmacokinetics describes what the body does with a drug after it is administered. The factors involved include how quickly the drug is absorbed and enters into the systemic circulation; the biological tissues and fluids where the drug is distributed within the body; how long the drug resides within the systemic circulation and tissue compartments; and how it is eliminated. These processes are generally described using the acronym ADME (Absorption, Distribution, Metabolism and Excretion). They determine the amount of drug that is ultimately available at the intended site of action. For a drug such as levofloxacin they determine how much drug is available to interact with a particular pathogen causing the infection.

Pharmacodynamics describes the interaction of the drug with the pathogen causing the infection. This interaction includes whether the drug leads to a bacteriostatic or bactericidal action; whether any persistent effects or subinhibitory anti-infective activity will be imparted upon the pathogen after the drug has been cleared from the infectious site; and whether the drug interacts directly or indirectly with the immune system to impart antimicrobial activity.

Bacteria can be differentiated into two main types, Gram-positive and Gram-negative, reflecting their appearance when treated with the Gram stain. They are responsible for a number of serious and unpleasant diseases. The most common community acquired infections in 1985 were respiratory tract infections, followed by urinary tract and skin and soft tissue infections. In the hospital, the most common infections were respiratory tract infections, urinary tract infections, infections of the blood and skin, and soft tissue infections. In all, respiratory tract infections, whether in the community or in hospital, were responsible for approximately 55-70% of all infections. S. pneumoniae, a Gram-positive bacterium was (and is) the leading cause of community acquired respiratory tract infections such as community acquired pneumonia and was responsible for the majority of deaths associated with that disease.

Urinary tract infections, the second most common type of infection, accounted for approximately 15% of all infections and were caused mainly by Gram-negative bacteria including, most importantly, E. coli and P. aeruginosa.

I think it important to have regard to the history of quinolone development because it reveals just what an active field of research it was and the wide variety of structures it yielded. The following is an outline of just some of the highlights and is derived largely from the evidence of Professor Wentland.

The first quinolone antibacterial agent was discovered by Sterling-Winthrop in the late 1950s, as a by-product of its work on malaria. It was called nalidixic acid and is depicted below:

It was only used to treat urinary tract infections caused by certain Gram-negative bacteria but was considered to be a major breakthrough in the antimicrobial field as it was not derived from natural product sources, but rather via chemical synthesis. As Professor Wentland observed, it presented chemists with the opportunity to make hundreds if not thousands of other analogues.

In 1966 scientists from Warner-Lambert reported an important new compound, later named oxolinic acid:

Relative to nalidixic acid, oxolinic acid displayed significantly greater antibacterial potency with a somewhat broadened spectrum to include some Gram-positive organisms. Notably, this compound, like nalidixic acid, contains a carboxyl group at the 3-position. This group was thought a stringent requirement for activity throughout the period to June 1985.

Over the course of the following few years many related compounds were developed, including, most notably, cinoxacin and miloxacin which, like oxolinic acid, had bridges at the 6,7 positions on the second quinolone ring.

The 1970s saw the synthesis of a number of important compounds, including flumequine and pipemidic acid:

Flumequine's tricyclic structure and antibacterial properties were widely known by researchers in the 1970s and 1980s. It was the first quinolone to recognise the benefits of fluorine substitution on the quinolone core structure. It had a broader spectrum in vitro activity than nalidixic acid, displaying some activity against Gram-negative bacteria (for example E. coli) but poor activity against Gram-positive bacteria (for example S. pneumoniae). However, its activity against E. coli along with its pharmacokinetics did create potential for it to be used for urinary tract infections and it was advanced to clinical trials. Although it never attained approval for human use in the US and is not approved in the UK, Professor Wentland believed it to be available for veterinary use, and in some places human use, in 1985. However, it was generally less potent in vitro than the later fluoroquinolones (for example, norfloxacin) and, by 1985, would not have been of much interest of a medicinal chemist interested in developing an improved quinolone, a point to which I must return in considering the allegation of obviousness.

The other structure depicted is pipemidic acid. This is characterized by an additional nitrogen in the core bicyclic ring and a piperazine ring (arrowed) in place of the methyl group. It was generally recognized that the piperazine ring at the 7-position was responsible for the improved in vivo antibacterial properties pipemidic acid displayed. It came to the market in 1975.

Other workers pursued the effect of substitutions at the 7-position. Among them were chemists at Sterling Winthrop who developed rosoxacin:

They found that a 4-pyridinyl group at the 7-position (arrowed) of a quinolone was responsible for improved potency relative to nalidixic acid.

Another breakthrough came in 1979 with the development by Kyorin scientists of norfloxacin:

This is a quinolone incorporating 6-fluoro and 7-piperazine groups in the same molecule. It was found to have more potent and broader spectrum antibacterial properties than any predecessor compounds. In particular, it exhibited enhanced activity against Gram-negative bacteria, including P. aeruginosa, and improved activity against some Gram-positive bacteria (but not including S. pneumoniae) compared to nalidixic acid. It also had good pharmacokinetics, having a half-life that allowed for twice daily dosing (400mg twice daily). However, its attainment of good systemic concentrations relative to its MIC values was poor. This limited its indications to the treatment of non-systemic infections such as those of the urinary tract, where it achieved sufficient urinary concentrations to be bacteriologically and clinically effective.

To try to address the problem of low serum concentrations, attempts were made to increase the daily dosage of norfloxacin to 1200mg or even 1600mg/day. However, its poor water solubility led to the occurrence of crystalluria (precipitation of the drug in the urine) in some patients. This side effect prevented increasing the daily dose of norfloxacin to produce a serum concentration high enough for the treatment of systemic infections.

Nevertheless, it made a big impact on the quinolone field with the promise of systemic activity due to its good oral antibacterial efficacy in animals and humans and few side effects. Norfloxacin was approved for use in humans in 1983. In response to its emergence, Sterling-Winthrop set up a specific task force for quinolones. It perceived what Professor Wentland described as an unlocked potential and each chemist began to make about two new targets per month.

Pefloxacin, a similar analogue, was also developed in 1979. In this case the nitrogen at the distal end of the piperazine ring is methylated:

It was found to display higher activity than norfloxacin when administered orally and was approved in France for human use in 1985.

Enoxacin, reported by Dainippon in 1980, rang the changes yet again with the introduction of a nitrogen at the 8-position:

This molecule proved to have better oral activity than norfloxacin and was approved in Japan for human use in 1986.

In 1982, Daiichi reported another very significant development, the production of ofloxacin. Its structure is shown in paragraph [2] of this judgment and it can be seen to have a 1,8-position bridge. The development of ofloxacin represented a significant improvement over norfloxacin including greater potency against both Gram-negative and Gram-positive organisms, better pharmacokinetics allowing for higher serum concentrations and facilitating the treatment of systemic infections and availability of both oral and intravenous dosage forms. Ofloxacin also appeared to be safe and, as I have mentioned, was approved for use in humans in 1985. However, I am satisfied it was still not regarded as a wholly satisfactory treatment of community acquired respiratory tract infections and, in particular, pneumonia.

At about the same time, Sterling-Winthrop’s development efforts led to new quinolones having 1-alkylamino groups. One of the most significant compounds was made by Professor Wentland and called amifloxacin:

He arrived at this structure because of his understanding (which he believed to be general) in the early 1980s that an ethyl or equivalent group at the 1-position of a quinolone was optimal for antibacterial activity. Amifloxacin was found to have exceptional in vivo qualities but was never marketed following Sterling-Winthrop’s change in direction in 1988.

This brings me to ciprofloxacin. The novel feature of the molecule is a cyclopropyl group at the 1-position:

It was first reported by Bayer in 1983 and found to have significantly improved activity over norfloxacin against a variety of Gram-negative organisms, including P. aeruginosa, and some Gram-positive organisms. This activity generated considerable excitement. Ciprofloxacin also displayed favourable serum pharmacokinetics that allowed for the treatment of systemic infections using 500mg twice daily dosing, with limited adverse effects. It improved upon norfloxacin in two major ways: its MICs were markedly lower than norfloxacin's and it achieved higher peak serum concentrations. Thus the ratio of serum concentration to MIC was greatly increased, allowing for the treatment of systemic infections. By comparison with ofloxacin, it was slightly more active against both Gram-negative and Gram-positive bacteria but it had a lower serum concentration and, once again, was not generally thought to be a complete answer to the problem of community acquired infections likely to have been caused by S. pneumoniae.

There was another feature of the disclosure that was also of great interest to researchers. Until this time quinolones were generally synthesised using a technique called the Gould-Jacobs reaction. But the scientists at Bayer developed a novel process called cycloaracylation which permitted the synthesis of quinolones with a wide variety of substituents at the 1-position, such as primary, secondary and tertiary alkyl, cycloalkyl, and primary, secondary and tertiary amino, aryl and heteroaryl groups.

It is interesting to note the impact of ciprofloxacin on Sterling-Winthrop. Professor Wentland first became aware of it in October 1983 at a conference at which he and his colleagues were due to announce amifloxacin. He immediately appreciated that ciprofloxacin was a superior molecule and described its disclosure as crushing. At the time Sterling-Winthrop was one of the leaders in the field. It had two quinolones on the market, namely nalidixic acid and rosoxacin but faced competition from Kyorin with norfloxacin, Daiichi with ofloxacin and now Bayer with ciprofloxacin. The pressure was to find new quinolones with as good or better profiles than these newer agents. It put another task force together under the leadership of Professor Wentland. There were times in the period up to June 1985 when up to 40 medicinal chemists at Sterling-Winthrop were designing or synthesising new quinolones, and particular attention was focused on ciprofloxacin and the new cycloaracylation process.

I am satisfied on the evidence that in 1985 ciprofloxacin was recognised to be the “class leader”. The quinolone field was extremely active with many groups seeking to identify further and different molecules which would improve on the pharmacokinetic and pharmacodynamic properties of ciprofloxacin, particularly against Gram-positive bacteria, such as S. pneumoniae, responsible for causing community acquired infections. There was also another factor active in the minds of researchers at that time. They were concerned about the development of resistance, as Professor Zhanel explained. This was a further driver to develop structurally diverse compounds.

Overall, it had been found that a variety of substituents, particularly at the 1 position, showed high activity. However, the relationship between structure and activity (the “SAR”) was not well understood and while this created a measure of uncertainty, it was appreciated that it left a wide choice of substituents open for further study, with the hope of developing a molecule of even greater activity than ciprofloxacin. As Professor Wentland explained, and I accept, it was a time of excitement and optimism.

There was no dispute that the skilled medicinal chemist would have known the basic principles of stereochemistry and, in particular, would have had a good understanding of isomers, stereoisomers, chirality and enantiomers. He would have known that one of the characteristics of enantiomers is that they rotate the plane of polarised light in opposite directions (hence dextro-rotatory (+) or levo-rotatory (-)), and so are examples of optically active substances. He would also have known that, this characteristic apart, the physical properties of enantiomers are generally the same. Similarly, he would have known that a racemate is a mixture of equal parts of enantiomers, that the rotation caused by one enantiomer is cancelled by the opposite rotation of the other and hence the racemate is optically inactive. Whether or not a compound is chiral is something that would have been immediately apparent.

As explained in the well known textbook, Organic Chemistry, by Morrison and Boyd, the nature of most physical properties of enantiomers means that they cannot be separated by ordinary methods such as fractional distillation, fractional crystallisation or chromatography. To obtain a pure enantiomer, the chemist therefore has to synthesise it using chiral starting materials or using chiral reagents or, alternatively, separate the enantiomers from the racemate (or at an earlier stage in the synthetic pathway) by using special resolution techniques such as those involving diastereomers. This is a topic to which I return below.

A good deal of evidence and time was directed to this issue. It was addressed by the experts and the parties both generally and specifically in relation to the field of quinolones.

In relation to the former, the dispute rather melted away. Dr Spargo and Dr Newton were consistent that the enantiomers of a chiral compound are likely to have different properties in vivo. It is not possible to make a prediction as to how different they will be but the best theoretical improvement in MIC is, at least in general, twofold. Dr Spargo also explained that one enantiomer may have better ADME properties than the other and, moreover, the skilled person would have expected the physical properties (including solubility) of the racemate to be different to that of its enantiomers. These opinions were supported by the fact that Glaxo resolved all the chiral molecules it had under development in the 1980s, including Zofran and Serevent.

Professors Wentland and Davies initially appeared to adopt a rather different line. But it became apparent there was, in the end, very little between them. Professor Wentland accepted that in many cases medicinal chemists studied the stereochemistry of drug molecules so they could gather information about the nature of the drug receptor interactions and receptor characteristics. However, he was consistent throughout his evidence that this did not occur to anyone in the team researching quinolones at Sterling-Winthrop, a matter which I develop in a moment. Professor Davies’s position was that biological systems are very complex and it is simply not possible to make any prediction as to whether the enantiomers will be equipotent, whether one will have essentially all the activity and the other none or whether the true position will be somewhere on the spectrum between these two extremes.

In so far as any difference remained between the parties’ respective positions at this general level, I prefer that of GUK. Many textbooks and papers, including Morrison and Boyd, were explored in evidence and they are consistent in revealing an appreciation by medicinal chemists in 1985 that the stereochemistry of a molecule can play a major role in its pharmacological properties. It was well understood that protein receptors are composed of a complex array of chiral amino acids and it follows that the drug’s binding site provides a chiral environment. In circumstances where the drug substance has an asymmetric carbon atom it was known that the binding efficiency of the individual enantiomers to a given receptor may be different. I return to consider how this relates to the position in the quinolone field as a separate topic a little later in this judgment.

Dr Newton explained in his first report that during the mid 1980s the FDA began to require information not only in relation to racemates but also in relation to their constituent enantiomers in order to evaluate the relative benefits and practicability of developing those individual enantiomers. He said he became aware of these issues as a result of his involvement in the development of Zofran and Serevent and felt confirmed in his view by the publication by the FDA in February 1987 of guidelines on acceptable approaches for meeting regulatory requirements. Although nearly two years after the priority date, they provide an insight into the thinking of the FDA at that time. In relation to a new drug application, they stated that impurities might have significant clinical or toxicological effects and that even in racemates, enantiomers might be considered as impurities. In relation to an investigational new drug (“IND”) application, they suggested, inter alia, that sponsors should ideally have either separated the enantiomers of a chiral molecule or synthesised them independently, that physical and chemical information about each should be provided or might be requested and that individual enantiomers might need to be studied for pharmacological and toxicological properties.

Further support for Dr Newton’s position came from a paper presented by Dr Kumkumian, the Assistant Director of Chemistry at the Office of Drug Research and Review at the Center for Drugs and Biologics of the FDA, in March 1986. This dealt with racemates at some length. In summary, Dr Kumkumian suggested that data submitted on substances that do or can exist as stereoisomers should include a discussion of the possible isomers which might result from the process of manufacture, and the results of any studies upon them. He recognised that the decision whether to market an isomer or a racemate was one primarily for the manufacturer but observed that it was very helpful for the FDA to be provided with information as to why a particular form had been chosen. In his conclusions, he observed that it was known that differences may exist between enantiomers and that it was becoming more routine for scientists to assess them. Absent full information on the biological properties of the enantiomers, decisions on the acceptability of the proposed product became more difficult for the FDA. He also noted that if the pure enantiomers could not be obtained or if neither had major advantages with respect to its pharmacological profile, the racemate might then be developed. But any information relating to the evaluation of the racemate and its individual enantiomers should be submitted, would assist the FDA and might expedite the review process.

Dr Newton’s evidence was answered by Dr Weissinger. I have summarised her background earlier in this judgment and she was very well placed to assist me as to the policy of the FDA in relation to stereoisomeric drugs in the mid 1980s. I am entirely satisfied in the light of her evidence that the FDA did not require pharmaceutical companies to assess or submit data concerning the relative activity or toxicology of the individual enantiomers of a racemic compound. What they were required to do was characterise the compound in terms of its identity, quality, purity, strength and stability. They had to identify any chiral centre and explain the form the drug the subject of the application took. If it was said to be a racemate then the optical rotation was required to confirm that was so. If they had details about the activity or toxicology of the individual enantiomers then they were required to disclose them. In short, they were required to disclose all the information they had. She also accepted that a number of pharmaceutical companies did resolve racemates and that in some instances this made the FDA’s job easier.

Overall, Dr Weissinger’s evidence was essentially consistent with that of Dr Newton subject to the qualification that the FDA did not require pharmaceutical companies to provide data on the individual enantiomers of a racemate. In this respect I accept the evidence of Dr Weissinger. I am also satisfied the evidence as a whole confirms that in 1985 the ordinary medicinal chemist would have been aware that the individual enantiomers of a racemate might well have different biological properties to each other and to the racemate itself.

Professor Davies expressed the opinion that the common view in 1985, and even today, was that resolution of racemates was a black art. By this I understood him to mean that resolution is an empirical process. It is simply not possible to make any prediction at all as to whether a particular technique will work. He also reviewed the specialist monographs and identified no fewer than 13 methods employed for the purpose of trying to separate individual enantiomers by 1985.

I accept the evidence of Professor Davies that it was not possible to predict with any degree of certainty whether and, if so, how the enantiomers of a racemate could be prepared. But I also have no doubt that a routine starting point for anyone seeking to resolve a racemate was to react it with an enantiomerically pure reagent that converted the individual enantiomers into diastereomers. As Morrison and Boyd explain, the majority of resolutions depend upon the reaction of organic acids with organic bases to yield diastereomeric salts. If, for example, a racemic acid is reacted with a base which is present in only one configuration, the result is crystals of two different salts which are neither superimposable, nor mirror images. These salts may have different physical properties, including solubility in a given solvent, and it may therefore be possible to separate them by, for example, fractional crystallisation. Once they have been separated, it may then be possible to recover the optically active acid from each salt by the addition of a strong mineral acid. Exactly the same approach can be applied to racemic bases.

If the compound has a suitable reactive centre then another generally known technique was to react it with a chiral reagent to form a pair of diastereomers via covalent bond formation. Again these might be separable by fractional crystallisation and the enantiomerically pure derivatives converted back into the enantiomers of interest using synthetic procedures that did not compromise the stereochemical integrity of the chiral carbon atom.

Another well known approach was to attempt to develop a total chiral synthesis of the drug. This could involve either starting from a commercially available chiral material or the use of chiral reagents to introduce chirality. Often this would involve a completely new synthesis and, as Dr Newton explained, it might not be the most attractive option.

If the synthetic route for making the racemate was known to involve a chiral intermediate then yet another possibility was to apply a separation technique to that intermediate and thereafter seek to form the desired enantiomer of the final product by one or more stereospecific reaction steps. This latter aspect is important. As in the case of the covalent separation technique, the synthetic procedures used in those reaction steps must preserve the stereochemistry of the chiral carbon atom. At least at the investigation stage, it was therefore a less attractive route than attempting to resolve the racemate.

As Professor Davies explained, and I accept, these were some of the methods available in 1985 for attempting to resolve a racemate. But they then break down into an almost infinite number of individual permutations of different reactants and conditions that might be tried. Even if the right method was chosen, the specific conditions adopted might not work. No resolution was (or is) simple and the skilled person would have known there was no guarantee it could be achieved.

A question arises as to whether the technique of preparative HPLC was common general knowledge. This is important because it was the technique used by GUK in its experiments in support of its case of obviousness. Dr Spargo explained in his first report that the technique was conventional in 1985 and, although Professor Davies made various criticisms of those experiments in his report, he did not suggest the technique was not conventional or generally known. The issue arose during the course of the cross examination of Dr Spargo when he accepted he had no direct experience of preparative HPLC at that time but suggested the answer might be apparent from the textbook Enantiomers, Racemates and Resolutions by Jacques and Collet, 1981 (“Jacques”). The authors say, at page 329, in dealing with the separation of covalent diastereomers: “…high pressure liquid chromatography for preparative purposes is now carried out at least as often as separation by crystallization”.

Much the same picture merges from another textbook, Chromatographic Separations by Souter, 1985. He explains that HPLC had been shown to be of great value in stereoisomer chromatography and could be used in several modes for stereoisomer separations. For example, he says, enantiomers may be separated on chiral chemically bonded phases or chiral additives may be mixed into the eluting solvent such that complexation with and consequent resolution of the enantiomeric solutes occurs. He shows that many different column materials and packings were available and that they had different uses and limitations. He also provides details of preparative or semi-preparative separations in his table 23, which he describes as being suitable for the reader’s use as a starting point for experiments.

Dr Newton also gave evidence that preparative HPLC was in general use but accepted in cross examination that it was a specialist technique and, importantly, that the availability and suitability of particular columns for particular purposes were matters outside his expertise. The evidence of Professor Davies under cross examination was broadly consistent. He too said preparative HPLC was a specialist technique and not something routinely used by most chemists.

In the light of this evidence I accept that preparative HPLC was not routinely used by medicinal chemists in 1985. But I believe it was a technique of which they would have been aware. They would have known of and been familiar with the use of HPLC for analytical purposes and they would also have known of its availability for use by specialists for preparative purposes. But to determine whether any particular column would work for preparative purposes and the conditions under which it should be used would likely need experimentation. Moreover, and as I shall explain, it has not been shown that the equipment used by GUK in the work up of its experiments was available, let alone generally known, in 1985.

There was a significant dispute about whether Jacques and its contents were common general knowledge in 1985, its importance being that it mentions menthyl chloroformate as a suitable reagent with which to react a racemic amine to produce diastereomers, this being the reagent used in GUK’s experiments.

In his reports, Dr Spargo described Jacques as a seminal text and stated that the skilled team would have been aware of it and members of the team would have referred to it as appropriate. He accepted it was specialist but said that the skilled team would refer to it as a matter of course when necessary. He recalled that a well thumbed copy was to hand during his time at Pfizer from 1988-2003. Under cross examination, he explained that he became aware of it when he needed to resolve a compound and found it in the library. He subsequently found it was also in some of the laboratories.

Professor Davies disagreed. He was aware of three specialist books on the subject of resolution that were available in 1985. These were (a) Jacques, (b) Optical Resolution Procedures for Chemical Compounds edited by Paul Newman and first published in 1981, and (c) Stereochemistry of Carbon Compounds by Eliel and first published in 1962. Because his particular area of expertise was the preparation of homochiral compounds, he had a copy of Jacques, to which he referred regularly, and he had access to Eliel. But he did not believe any of these were the sort of textbook the skilled medicinal chemist would have had to hand in 1985. Under cross examination, Professor Davies explained that Jacques is a very advanced textbook but does have the advantage that it contains a significant number of examples. Nevertheless, he did not think many copies had been sold and could not recall ever having seen it on walking round many laboratories and pharmaceutical companies at that time. However, he accepted it contained descriptions of resolutions that had worked in the past although he suggested it was primarily concerned with optimisation of resolutions rather than providing initial solutions. He also recognised it was referred to by way of footnotes in Advanced Organic Chemistry, 3^rd edn., a textbook by March published in 1985. But this he considered to be an advanced text, albeit one used regularly by chemists in industry.

Finally I should mention Dr Newton and Professor Wentland. Neither had apparently come across Jacques although it appears Dr Newton may have had a copy of Eliel.

In the light of all this evidence I am not persuaded that Jacques was common general knowledge in 1985. I am satisfied it was one of a number of specialised source or reference books available in some but not all pharmaceutical companies. I have no doubt it would have been known to some specialists in chiral chemistry and it might have been found by others interested in resolving a racemate. But it has not been established that the skilled team would have known of its existence or referred to it as a matter of course either directly or through a specialist.

Professor Wentland had a very clear view as to the mindset of those in the quinolone field in 1985 and as to the approaches they took to the development of new compounds. It was his experience that they were influenced by the poorly understood SAR and the difficulty of predicting whether a new compound would have desirable characteristics. As a result, they might pick combinations of known and novel groups and append them to a particular known quinolone core. Groups would be chosen on the basis they were known to be beneficial for good activity or were analogues of such groups. A common strategy was to introduce one novel group while keeping other known groups constant. Another strategy was to try and design novel core structures to which known or novel groups could then be appended. Overall, the difficulty of making predictions and the scope for variation and the potential for major improvements made it a fertile area of research. Indeed, it was so absorbing that Professor Wentland could not recall anyone at Sterling-Winthrop even suggesting resolving a racemic quinolone, despite several having been made. As he put it, they were not thinking of that; it was not something on their radar screen. They had a focussed objective which was to find a new chemical entity which had a certain antibacterial and ADME profile and to get there they were looking at new appendages, new core structures, and combinations of each. He also thought that given the poorly understood SAR and belief that binding took place according to the induced fit model, there was reason to suppose that the target had considerable flexibility and could accommodate a number of differently shaped molecules. This made it even less attractive to set about what might be a very difficult resolution.

Dr Spargo and Dr Newton found the evidence of Professor Wentland very hard to understand. For them it was simply a matter of the interactions of biological systems and there was no reason to treat the quinolones any differently to any other class of enzyme inhibitors. However, it must be acknowledged that neither was in the quinolone field at the priority date and, as Dr Spargo fairly accepted, his position was inconsistent with the absence of any mention of stereochemistry in the major review articles of the time.

I must also take into account the evidence that has now emerged that a number of teams were in fact working on enantiomers of chiral quinolones before and after the priority date. First there were of course the inventors at Daiichi, who had earlier published the discovery of ofloxacin. Then there is the work of the team at Riker, including Dr Gerster and Dr Rohlfing, who investigated the enantiomers of flumequine from 1982. Shortly after the priority date, in December 1985, workers at Bayer filed the Bayer application relied upon as a novelty citation and which discloses the application of the method developed by Bayer in 1983 to synthesise ciprofloxacin to the production of levofloxacin. At about the same time, Professor Mitscher and co-workers at Abbott also began work to develop a process for the production of the individual enantiomers of ofloxacin and duly filed a patent application in April 1986. Interestingly, they noted in an article published later that year, J Med Chem 1986, Vol.29, 2044-2047, chiral preferences of drug candidates was a subject of considerable contemporary interest but one that had rarely been examined as yet amongst the quinolone antimicrobial agents. Roche too appear to have produced data on individual enantiomers of one of their compounds, which they presented in 1985. Finally, I was referred to work by Kyorin which appears to have been conducted in or shortly before March 1986 when it was introduced into one of their patent applications; but since this was after the date of the publication by Daiichi of their work on levofloxacin, I think it merits little weight.

Overall, I think a relatively clear picture emerges. The quinolone field was unusual in that workers recognised the need for and perceived an opportunity to discover new chemical entities of ever greater efficacy. The discovery of norfloxacin, ofloxacin and ciprofloxacin put great pressure on researchers to identify new quinolones having even better bacterial profiles. It was here their energies were primarily directed. Hundreds of chemists were making new compounds each year and, not surprisingly, they would opt not to make a compound if its synthesis was difficult. Compounds which were difficult to make would get a low priority in the laboratory. This was not an environment conducive to the investigation of stereochemistry. Resolution of a racemate might result in a twofold increase in activity, at best, and it could be a good deal worse. It was quite possible that activity might lie more in one enantiomer than the other, but not greatly so. Moreover, resolution might prove very difficult to achieve. As a result, for many chemists, including those at Sterling-Winthrop, I am satisfied that resolution of racemates was something that would simply not have occurred to them at all. For others, it plainly did. But it was not a routine path to follow and, for the ordinary chemist, I believe it was something which he might well not have considered and, if he did, then it would not have been a high priority, absent some particular reason for doing so. I return to consider the position in relation to ofloxacin when I address the allegations of obviousness.

Priority is important in this case because it is accepted by Daiichi that if the first two priority dates are lost, then the Bayer application anticipates all the claims in issue.

The argument advanced by GUK runs as follows. The Patent contains subject matter (at page 3, lines 13-18 and page 12, line 47 to page 13, line 35) concerning reduced toxicity and increased water solubility which is not contained in either of the first or second priority documents (respectively filed on 20 June 1985 and 11 October 1985). This matter is not entitled to a priority date earlier than the filing date of the Patent (20 June 1986) or the filing date of the third priority document (28 January 1986). Accordingly, if and to the extent Daiichi rely upon any of this matter as supporting the inventions of any of the claims of the Patent then those inventions are not entitled to claim the priority dates of the first or second priority documents.

In considering this contention it is, in my judgment, important to distinguish between the priority date of a claimed invention and the priority date of matter contained in an application. An invention in a patent application is entitled to the date of an earlier priority application in the circumstances provided for in section 5(2)(a) of the Patents Act 1977. In short, the question is whether the invention is ‘supported by matter disclosed’ in the earlier application. Support in this context comprises two requirements. The first is that the earlier application enables the invention: Asahi Kasei Kogyo Application [1991] RPC 485 (HL). The second is essentially one of disclosure and was explained by the Enlarged Board of Appeal of the EPO in G02/98, Same Invention, [2001] OJ EPO 413; [2002] EPOR 167:

"The requirement for claiming priority of 'the same invention', referred to in Article 87(1) EPC, means that priority of a previous application in respect of a claim in a European patent application in accordance with Article 88 EPC is to be acknowledged only if the skilled person can derive the subject-matter of the claim directly and unambiguously, using common general knowledge, from the previous application as a whole."

This second requirement was further considered by the Court of Appeal in Unilin Beheer v Berry Floor [2004] EWCA (Civ) 1021; [2005] FSR 6 at [47]-[50]:

"47.

Mr Tappin submitted that the position was simpler. He submitted that priority can be accorded to any invention the subject-matter of which is disclosed in and enabled by the priority document. Any claim or "consistory clause" (jargon for a repetition of the main claim in the body of the specification) in the priority document is not determinative. The important thing to focus on is whether the priority document as a whole discloses matter (i.e. information) which effectively gives the skilled man what is in the claim whose priority is in question and that that information is enabling.

48.

I think Mr Tappin must be right. That is what both Biogen and G02/98 are saying. The approach is not formulaic: priority is a question about technical disclosure, explicit or implicit. Is there enough in the priority document to give the skilled man essentially the same information as forms the subject of the claim and enables him to work the invention in accordance with that claim.

49.

Before going to the details of the priority document in this case I should deal with Mr Carr's submission about the main claim or consistory clause of the priority document, i.e. that although not determinative it is nearly so. That he could not get out of GO2/98 or indeed any other authority. GO2/98 refers to "the previous application as a whole," not the main claim nor the "main statement of invention" nor the "consistory clause". Likewise there is nothing in Art.87 which compels or suggests this conclusion. And Art.4H of the Paris Convention is positively against it. The claims (if any—there is no rule that there should be) of the priority document are not determinative. They are just part of its disclosure. For the purposes of priority one just looks at the disclosure as a whole.

50.

If the rule were otherwise one of the main functions of a priority document would be lost. Inventors and their advisors would have to start worrying not only about the technical information disclosed in the document but how it was to be claimed: have I drafted my main claim or consistory clause broadly enough? That is not the purpose of the system: the purpose at this point is to get the information justifying the later claim into a patent office of a Union country. If you do that, you can have your priority, whether you express that in a proposed claim, consistory clause, statement of invention, other text or drawing or in any combination of these. Time is of the essence because the world-wide system (except for the Americans) works on the first to file basis. The detailed framing of a claim based on that information may then be done within the Convention year."

Application of this test in the context of the present case can yield only one answer. Each of the claims in issue is plainly disclosed in the first and second priority documents. Moreover, there is no dispute that those priority documents do permit the skilled person to perform each of those claimed inventions without undue difficulty. In my judgment the fact that those priority documents contain nothing about reduced toxicity and increased water solubility is irrelevant. They are not features of the claimed inventions.

A separate and distinct question is whether it is permissible for Daiichi to rely upon the disclosure of reduced toxicity and increased solubility in support of their case of non obviousness. For the reasons I explained in Generics Lundbeck [2007] EWHC 1040 (Pat) at [229]-[235], I do not believe it can. After referring to Richardson-Vicks Inc’s Patent [1995] RPC 568 and Glaxo Group Ltd’s Patent [2004] RPC 43, I concluded at [235]:

“235 Both of these cases were concerned with synergy. But it seems to me that the logic behind them is not limited to such cases. A patentee cannot seek to bolster the inventive nature of his monopoly by relying on a discovery which he had not made at the time of the patent. That is the position here. At the date of the Patent, Lundbeck had not found that escitalopram was more efficacious or was effective in treating more patients than citalopram. Those discoveries were not made until some time later. They are nowhere hinted at in the specification and could not have been predicted from what is described. In these circumstances I do not believe that it is legitimate for Lundbeck to rely upon them in support of the alleged invention.”

In my judgment the position must be the same in relation to an invention which claims priority from an earlier application. Where the claim to priority is properly made, as here, I do not believe a patentee can seek to bolster the inventive nature of his monopoly by seeking to rely on a discovery which he has not made at the date of the priority document and which is not foreshadowed in his disclosure and could not be predicted from what he has described.

There are two separate anticipation attacks. The first is based upon the Bayer application and depends upon the priority claim. I have concluded that the Patent is entitled to its first claimed priority date and so the attack fails. The second is based upon the Daiichi prior art, that is to say, the 005 Patent and Drugs of the Future, to which I now turn.

It is accepted that both the 005 Patent and Drugs of the Future disclose ofloxacin and a way of making it. Moreover, both documents disclose the use of N-methylpiperazine in the final reaction step. GUK says that the skilled person would immediately recognise that ofloxacin is a racemic mixture of two enantiomers and that accordingly levofloxacin is disclosed, as is the process of claim 5.

In support of this contention GUK relies upon a series of decisions of the German Federal Patent Court, the Bundespatentgericht. My attention was drawn to two decisions in particular. In Neolab v Lundbeck, a decision of August 2007, the court put it this way at II.1.:

“A chemical compound with an asymmetric carbon atom is no longer novel in the form of one of its enantiomers if the skilled person’s attention is concretely drawn in a prior publication to the enantiomer and if said person is able to prepare the compound by virtue of this concrete instruction and his general expert knowledge. It is not necessary that the compound was already prepared in reality… The novelty of the enantiomers of a chemical compound or the novelty of the enantiomers of a compound group described, for example, on the basis of a Markush formula must, however, also be denied if a prior publication just describes the preparation and the chemical structure of the compound or group of compounds occurring in the form of an enantiomer mixture, specifically as a racemate, without the existence of the corresponding enantiomer as a substance being pointed out expressis verbis. In such a case there is no room for the acknowledgment of the novelty of an individual enantiomer if the skilled person recognizes its presence in the form of an enantiomer mixture, thus readily implies the individual enantiomer and can also obtain the same singly by applying conventional separating methods and by taking reasonable efforts….”

It can be seen the court took the view that if the skilled person would recognise that the disclosed compound comprised a racemate and so contained a mixture of two enantiomers and if he could resolve the mixture using conventional techniques then the disclosure would deprive each enantiomer of novelty.

The court applied the same approach in Basics v Warner Lambert, a decision of October 2007, where it said at II.1.a):

“According to Federal Patent Court practice ….. the disclosure of a document pertinent as prior art is not limited to a literal description, but comprises everything that is obvious to the skilled person or that a skilled person would add as essential or that he would read as implied when carefully studying the document. These principles that are related to the disclosure of a prior publication in the field of mechanics are also applicable in the field of substance chemistry under the proviso that the novelty of a chemical compound is to be regarded as anticipated if the skilled person will derive from a prior publication or a publication with an earlier priority a clear indication to the specific compound, i.e. if he will read the presence of this compound as being implied without difficulty, and if due to this indication he will be capable of obtaining said substance. It is not required that this substance has de facto already been produced. All that is required is the mere possibility of its producibility and, thus, its accessibility ….

What is characteristic in the present case of a stereoisomer is that, as a rule, compounds with one or more asymmetric carbon atoms in a chemical synthesis – other than natural substances in a biosynthesis – are present in a mixture of their stereoisomers with a more or less high portion of individual isomers. Therefore, the individual stereoisomers are already present in the reaction product. If a document pertinent as prior art discloses the preparation of a chemical compound with one or more asymmetric carbon atoms by way of a non-stereospecific reaction, this will be immediately evident to the skilled reader without requiring a further indication or even explicit naming of individual stereoisomers. Therefore, novelty of a stereoisomer (epimer, enantiomer, diastereoisomer) is to be denied if it is obvious to the skilled reader that it has already been disclosed in the prior art in form of its stereoisomeric mixture and, thus, is at his disposal by common separation processes, i.e. without difficulty ….. In that case, neither an indication nor an explicit naming of the respective stereoisomer is required, nor the naming or description of a working method of its isolation …..”

Interestingly, the Bundespatentgericht has recognised its approach is out of step with the EPO, as revealed by the following passage a little later in the decision at II.1.e)-f):

“…In consideration of Federal Court practice with regard to the novelty of chemical compounds …., the Senate cannot agree with these evaluations [by the EPO] that are based on a very narrow – sometimes blatantly referred to as “photographic” – notion of novelty [citing T181/82, T7/86, T286/87, T81/85], which is, by the way, not applied consistently even by the European Patent Office [citing T12/81, T12/90]. …

f)

The Senate cannot agree with Defendant’s point of view according to which the patentability of the patent-in-suit may not be assessed deviating from the Decision of the Board of Appeal of the EPO dated July 20, 2000 and from the Examining Division’s opinion dated October 19, 2006 for the sake of the required harmonization of national and European patent laws by a preferably uniform application of law ….. The Federal Patent Court emphasized this obligation. Irrespective of the known diverging practice of assessment in the individual contracting states, there are no binding rules of interpretation for novelty and inventive step comparable to those decided for the determination of the scope of protection of a European patent …… in the protocol for the interpretation of Art. 69 EPC. Moreover, it can also not be stated that, with regard to the presently relevant issue of novelty, in the majority of the EPC contracting states an application of law complying with the standard of valuation of the European Patent Office had become common practice that would form the basis of a decision-making according to Art. 31, sec. 3b of the relevant principles of interpretation of the Vienna Convention on the Law of Treaties of May 23, 1969 …”

The difficulty I have with the German approach is that it appears to conflate the issues of novelty and obviousness. The test that I must apply in considering the issue of novelty was explained by the House of Lords in Synthon BV v SmithKline Beecham [2005] UKHL 59; [2006] RPC 10. It has two requirements, prior disclosure and enablement but, for present purposes, it is the disclosure requirement which is of particular relevance. To satisfy this requirement, it must be shown that the prior art contains a clear description of, or clear instructions to do or make, something that would infringe the patentee's claim if carried out after the grant of the patentee's patent. As Lord Hoffmann said at [22]:

“If I may summarise the effect of these two well-known statements, the matter relied upon as prior art must disclose subject-matter which, if performed, would necessarily result in an infringement of the patent.”

A little later he made the same point at [24]:

“Although it is sometimes said that there are two forms of anticipatory disclosure: a disclosure of the patented invention itself and a disclosure of an invention which, if performed, would necessarily infringe the patented invention (see, for example, Laddie J. in Inhale Therapeutic Systems Inc v Quadrant Healthcare Plc [2002] R.P.C. 21 at [43]) they are both aspects of a single principle, namely that anticipation requires prior disclosure of subject-matter which, when performed, must necessarily infringe the patented invention.”

In the present case I think it is clear that this requirement is not satisfied. The 005 Patent and Drugs of the Future are both concerned with the ofloxacin racemate. Neither teaches or even suggests resolution of the racemate into its two enantiomers. Performance of the subject matter of these publications would result in the production of ofloxacin, not levofloxacin. It would not constitute an infringement of the claimed invention because, as GUK accepts, the claims do not extend to the racemate. I feel reinforced in this conclusion by the decision of the Court of Appeal in Generics v Lundbeck [2008] EWCA Civ 311. Although the point was not argued, Lord Hoffmann noted with apparent approval at [9] what he described as the settled jurisprudence of the EPO that disclosure of a racemate does not in itself amount to disclosure of each of its enantiomers. Similarly, Jacob LJ approached the matter as one of construction. At [50] he explained the question was whether the claim in issue covered the single enantiomer when in the racemate. In his opinion it obviously did not because the patentee was plainly not intending to cover the racemate. Just so here.

For these reasons the anticipation case based upon the 005 Patent and Drugs of the Future fails. I say nothing at this point about the requirement of enablement save to note that the ease or otherwise of resolution is a topic to which I must return in considering the allegation of obviousness.

It is convenient to consider the question of obviousness by using the structured approach explained by the Court of Appeal in Pozzoli v BDMO [2007] EWCA Civ 588. This involves the following steps:

It is of course the last question which is critical and requires the court to take into account all the relevant evidence. In doing so it is helpful to consider the nature of the obviousness case. Here it is, at least in part, that it was obvious at the priority date to seek to produce the enantiomers of ofloxacin and the skilled person could do so without invention. This is sometimes called an ‘obvious to try’ case, and it is one of the kind recently considered by the House of Lords in Conor v Angiotech [2008] UKHL 49. There Lord Hoffmann said at [42]:

“In the Court of Appeal, Jacob LJ dealt comprehensively with the question of when an invention could be considered obvious on the ground that it was obvious to try. He correctly summarised the authorities, starting with the judgment of Diplock LJ in Johns-Manville Corporation's Patent [1967] RPC 479, by saying that the notion of something being obvious to try was useful only in a case in which there was a fair expectation of success. How much of an expectation would be needed depended upon the particular facts of the case. As Kitchin J said in Generics (UK) Ltd v H Lundbeck A/S [2007] RPC 32 , para 72:

“The question of obviousness must be considered on the facts of each case. The court must consider the weight to be attached to any particular factor in the light of all the relevant circumstances. These may include such matters as the motive to find a solution to the problem the patent addresses, the number and extent of the possible avenues of research, the effort involved in pursuing them and the expectation of success.””

That is the approach I propose to adopt in this case.

I have addressed the skilled person and the common general knowledge and need say no more about them at this stage. The inventive concepts of the claims are straightforward. In the case of claim 2, it is levofloxacin. In the case of claim 5, it is a process for making levofloxacin which involves adding N-methyl piperazine to the intermediate denoted (V) in the Patent. I have considered the disclosure of the Daiichi prior art at paragraph [119] above. Having rejected GUK’s contentions on anticipation, it can be seen the differences between the Daiichi prior art and the claims in issue are that the Daiichi prior art discloses only racemic ofloxacin and a synthetic method for making it, whereas the claims are to the single enantiomer levofloxacin and a method for making that.

In summary, GUK contends it was obvious to seek to prepare the enantiomers of ofloxacin in the expectation that one of them would be likely to have better properties than the racemate. The argument then proceeds as follows:

Daiichi contests virtually every step of this path. It disputes it was obvious to seek to prepare the enantiomers of ofloxacin at all and it disputes it was obvious how to carry out the preparation, on the assumption the skilled person was minded to try. It also emphasises, rightly in my view, that obviousness is at the end of the day a single issue to be determined in the light of all these matters taken together.

I start then by considering whether there was any general motive to produce the enantiomers of ofloxacin in 1985. This must be considered against the background understanding of chirality that I have discussed at paragraphs [82]-[106] and the rather unusual nature of the quinolone field as I have described it at paragraphs [54]-[81] and [107]-[110].

There can be no doubt that as of 1985 ciprofloxacin was perceived to be the quinolone class leader. Nevertheless, ofloxacin was also recognised to be a clinically useful drug. Professor Zhanel described it as ‘very good’ with the ‘best pharmacokinetics of any quinolone’. It had very high serum concentrations, higher indeed than ciprofloxacin, bioavailability of almost 100%, very low protein binding and high concentrations in the urine on excretion. It was in Phase II clinical trials and looked extremely promising with excellent documented safety and efficacy. It was, in Professor Zhanel’s view, a drug which a pharmaceutical company would think was well worth having in its portfolio and it was seen as valuable for treating a range of infections. Against this background it was, perhaps, hardly surprising that Professor Zhanel also agreed that a drug which was more active than ofloxacin but similar in other respects would also be regarded as clinically useful in 1985 and would be worth trying for.

There is a further factor in GUK’s favour. As Dr Spargo explained, the advantage of starting with ofloxacin was that, as a drug in the clinic, it had already cleared many hurdles. As such, it was a promising base from which to ask: does chirality make a difference? By contrast, a chemical change would produce a new molecule which would require investigation from scratch.

Professor Wentland had a very different perspective to that of Dr Spargo. He was clear that workers were looking for more than a twofold increase in activity, the theoretical maximum that would be obtained if all the activity lay in one enantiomer. He also explained that in 1983 he was given the task of bringing the group at Sterling-Winthrop up to speed on quinolones. In doing he presented flumequine and ofloxacin as part of the history and noted they had centres of chirality. But he did not comment on their individual enantiomers because he had no knowledge of them and he did not discuss resolution.

Professor Zhanel was also firmly of the view that the ordinary skilled person would not have been interested in the enantiomers of ofloxacin. His perception was that it was doubtful that a single enantiomer, even if more active, would have a better pharmacological profile than the racemate and there was always a risk of greater toxicity. Time and energy would be better spent in trying to develop a new molecule.

I have to confess I have not found it easy to reconcile these two conflicting positions. Weighing the evidence as a whole against the background of the common general knowledge, I have reached the conclusion that by 1985 the skilled person would have been aware of the particular promise of ofloxacin as a pharmaceutical and its chiral nature. It was possible one enantiomer might have more activity than the other and that it would retain the other beneficial qualities of the racemate. Accordingly I believe he would have considered it worthwhile exploring whether ofloxacin could be resolved, but only to a point. I do not believe it was a goal that it was obvious to pursue relentlessly. The benefit in terms of activity might be relatively limited and it would have been recognised that resolution might be difficult or impossible to achieve. Moreover, the pharmacokinetics of a more active enantiomer would be unlikely to be better than those of the racemate and it might well be more toxic. It would also be susceptible to the development of resistance. By contrast, new molecules had the potential to be a good deal more active than ofloxacin and ciprofloxacin and I am satisfied the primary goal of those in the industry was to find them. Accordingly, I think it was obvious to investigate whether the enantiomers of ofloxacin could be separated relatively easily. If they could not, then I believe the skilled person would have redirected his efforts elsewhere, just as Professor Wentland explained workers would tend to avoid molecules they thought would be difficult to make. There were many other, potentially much more fruitful, avenues to explore.

That brings me to the second part of the analysis, namely whether the resolution could be achieved without invention. GUK contends that an obvious target for resolution was the intermediate denoted (V) in Drugs of the Future. The problem with this is that all the experts agreed that the obvious starting point would have been ofloxacin itself. Dr Spargo accepted this would have been the skilled person’s ‘first port of call’. Dr Newton agreed with Professor Davies that if the final product was available and could be made then typically the skilled person would try and resolve that first. However, I have no evidence that ofloxacin could be resolved without undue difficulty in 1985. So, if the skilled person set off down this path, I cannot assume he would have been successful.

On the assumption that his efforts to resolve ofloxacin were unsuccessful, the skilled person would have been presented with something of a dilemma. Should he continue and attempt to resolve one of the intermediates and, if so, which one? I am not persuaded that it was obvious to proceed beyond this stage and I think the skilled person would have concluded, as did Professor Wentland, that he was better off searching for new molecules. However, if he decided to continue then Professor Davies thought he would simply move back one or two steps to intermediates (IX) or (VIII). The reason for this is simple. The closer the intermediate is to the final product the less the risk of re-racemisation occurring when the resolved intermediate (assuming it has been resolved) is taken through the further reaction steps. Indeed, he expressed surprise that resolved intermediate (V) could be taken through to the individual enantiomers of ofloxacin without the loss of stereochemistry because of the nature of the intervening reaction steps.

Dr Spargo and Dr Newton preferred intermediate (V). This is the earliest chiral intermediate that has a ‘handle’ for chiral separation by the use of diastereomeric salts and covalent derivatives. It appeared to Dr Spargo to have two particularly attractive features. First, he thought it a planar molecule with relatively little conformational flexibility, which might confer a greater propensity for rigid solid state structures, perhaps increasing the chances that diastereomeric derivatives of (V) might exist in crystalline form – so permitting their separation. Secondly, the nitrogen atom at which the chiral derivative would be formed is in close proximity to the chiral centre, a property which he considered potentially advantageous for the separation of enantiomers. I have no doubt that these views were genuinely held but they did not find favour with Professor Davies. He believed it to be completely unpredictable whether a molecule will crystallise or not, depending as it does upon the thermodynamics and kinetics of the system. Nor did he consider it helpful to consider the location of the stereogenic centres.

Professor Davies also thought that, on the assumption a decision to resolve an intermediate had been taken, and that the skilled person was going to try to form a covalent derivative, he would not choose intermediate (V) because it is an aniline and hence a poor nucleophile and a weak base. In addition, the nitrogen atom is sterically hindered. Nucleophilicity decreases with increasing steric hindrance and both facts make it more difficult for it to react with a chosen resolving agent.

I have formed the view that all of the opinions expressed were reasonable and confirm the difficulty of making any sound predictions in this area of chemistry. The skilled person might just as easily pick intermediate (VIII) or (IX) as (V). Which one he chose would reflect his personal preferences and intuition. However, as in the case of ofloxacin, I have no evidence to suggest a resolution of any intermediate other than (V) would have been successful.

On the assumption the skilled person chose intermediate (V), GUK’s case is, in summary, as follows. Diastereomeric salts would be the first route for the skilled person to try, and for that purpose he would be likely to use a number of common chiral acids. For each acid the chemist would try to crystallise the product using a range of common organic solvents and solvent mixtures. If unsuccessful, the chemist would try other chiral acids or solvent systems and, at some point, would consider the covalent diastereomer approach. For an amine such as intermediate (V), menthyl chloroformate and Mosher’s acid would have been examples of the kind of agent worth trying .

At this point I must introduce the experiments. After reading only the 005 Patent and the Drugs of the Future Entry, Dr Spargo suggested to GUK’s solicitors that resolution of intermediate (V) in the Drugs of the Future Entry would form the basis of a preferred route to the enantiomers of ofloxacin. Those solicitors then asked him to suggest an experimental approach for resolving intermediate (V) using diastereomeric salts that would have been used by the skilled chemist as at June 1985 and he duly produced a protocol which was in evidence before me.

Apparently in accordance with Dr Spargo’s recommendations, a team of chemists at the University of Newcastle under Professor Roger Griffin attempted selective crystallisation of diastereomeric salts of intermediate (V) using a number of different resolving agents and solvent systems. However, a reproducible, scaleable resolution was not achieved with the conditions attempted.

Dr Spargo explained to me that further resolving agents, solvents and solvent mixtures could have been used, as he had contemplated in his protocol. However, he was told by GUK’s solicitors that, after that initial lack of success with chiral acids, Professor Griffin suggested that a covalent diastereomer approach be attempted. This Dr Spargo believed to be in accordance with his own view that a covalent diastereomer approach would have been attempted if the diastereomeric salt approach had not been successful after a reasonable time.

At that point he was asked by GUK’s solicitors to suggest a derivatising agent to form covalent diastereomers of intermediate (V). In response he suggested menthyl chloroformate and Mosher’s acid and he did so in the circumstances I shall elaborate in more detail in a moment. Dr Spargo was given to understand by GUK’s solicitors that Professor Griffin had independently also suggested the same two derivatising agents.

In the meantime, Professor Griffin and his team had carried out the experimental work that formed the basis of the Notice of Experiments. These describe the covalent derivatisation of intermediate (V) using menthyl chloroformate, the separation of the diastereomers using reverse phase HPLC and the liberation and characterisation of the enantiomers of intermediate (V). Initial attempts to separate the covalent diastereomers using thin layer chromatography (TLC) and selective crystallisation were unsuccessful and so the experimenters developed a reverse phase HPLC separation method. In due course, the experiments were repeated and the products of all the experiments were subjected to NMR, chiral HPLC and polarimetry analysis.

Dr Spargo also explained that the results in the Notice of Experiments and those obtained during the repeats show that the enantiomers of intermediate (V) are separable using this covalent diastereomer approach. The chiral HPLC analysis of the products obtained during the repeats shows >99.8% enantiomeric purity for both enantiomers. In the Notice of Experiments, lower levels of enantiomeric purity were obtained (about 95% and 91% for the (+) and (-) enantiomers respectively). He accepted that the polarimetry results show an inexplicable lack of consistency, but maintained that the chiral HPLC data demonstrate unambiguously the levels of enantiomeric purity obtained.

Daiichi level a number of criticisms at these experiments, some of which I believe are well founded and some less so. The first and most general criticism is that no one from the Newcastle team gave evidence before me. The experiments were, it is said, contrived litigation chemistry. In this regard reliance is placed upon the observations of Jacob LJ in SKB v Apotex [2005] FSR 23 at [77]-[78] and Aldous LJ in BASF v SmithKline Beecham plc [2003] EWCA Civ 872 at [75]-[76]. I consider this criticism too severe. I would be reluctant to impose on parties to patent litigation any general obligation which would result in an increase in the already substantial costs which such litigation entails. Often, experiments are relatively simple and transparent and are simply adduced as evidence that, when performed as described, particular results are obtained. However, this is not always the case and it must be understood that if relevant questions do arise as to how experiments were designed and how they came to be conducted as they were, and if the witnesses attending court are unable to address such questions, then the weight which the court can attach to the experiments may be substantially reduced.

Next it is said, rightly, that the experimenters at Newcastle did not work their way through the whole of Dr Spargo’s protocol in relation to diastereomeric salts. Over a period of two months they tried six different chiral acids, five different solvent systems, three different scales of reaction, four different volumes of solvents and a variety of different ways of promoting crystal formation. However, they did not try those different solvents with all the chiral acids; nor did they carry out any work with a number of other acids and solvents Dr Spargo had suggested. Professor Davies considered the skilled person would have continued with the salt approach much longer before attempting any other method of resolution. I accept that is so. However, I agree with GUK that the question is not whether the skilled person could have done other things with diastereomeric salts, but whether what was done with covalent derivatives was reasonable and obvious. In so far as it depends upon the extent to which the skilled person would have persisted with the salt approach then I must take it into account when making my overall assessment.

Daiichi then turns to the choice of menthyl chloroformate. The way in which this came to be suggested by Dr Spargo emerged during the course of the trial after GUK’s solicitors disclosed some attendance notes. Dr Spargo had previously indicated to GUK’s solicitors that Jacques was ‘the Bible’ and that was where the skilled person looking at resolution would go. When he was asked by those solicitors on the telephone what he would suggest for the covalent route, he did not have Jacques with him and so they read to him a list of the compounds which Jacques disclosed as having been used to resolve amines in this way, including menthyl chloroformate. At this stage they knew menthyl chloroformate had already been adopted by Professor Griffin. Interestingly, both Dr Spargo and Professor Griffin also suggested the use of Mosher’s acid, an expensive NMR shift reagent which is not mentioned in Jacques in connection with amines (and which, incidentally, does not apparently permit resolution of intermediate (V)). Not surprisingly, the disclosure of these conversations aroused a good deal of suspicion and has led Daiichi to suggest that I should attach little or no weight to Dr Spargo’s evidence and that the answer was ‘put under his nose’.

As I have said, I have no doubt Dr Spargo did not at any time attempt to mislead the court. I found him to be an entirely straightforward and honest witness. However, the difficulty I have with the evidence as it emerged is that Dr Spargo did not find the answer for himself. More fundamentally, I have found that Jacques was only one of a number of specialist texts to which the skilled person might have turned for assistance and its contents were not common general knowledge. If, for example, the skilled person had turned to Newman or Eliel it is simply not established he would have settled on menthyl chloroformate as a possible reagent at all. In particular, Newman shows that by 1981 approximately 1,000 amines had been resolved requiring 120 different resolving agents.

Much the same criticism is made by Daiichi of the choice of preparative HPLC to separate the diastereomers. I have no evidence as to how it came to be selected by Professor Griffin. However, it again emerged during the course of the trial that Dr Spargo had suggested it in the course of a telephone conversation with GUK’s solicitors along with flash chromatography, fractional crystallisation and reverse phase TLC. Indeed, in relation to preparative HPLC, he indicated this was ‘not as well developed in the 1980s’. I have addressed the issue of whether or not preparative HPLC was a technique which formed part of the common general knowledge in paragraphs [98]-[101] of this judgment and I have concluded that the skilled person would have known of its availability for use by specialists. But I have very little evidence upon which to conclude it was a technique to which he would have turned if flash chromatography, fractional crystallisation and reverse phase TLC were unsuccessful. GUK’s difficulties are compounded by the fact that Professor Griffin’s team carried out initial scouting work on modern columns. Indeed, it seems they optimised the separation on modern columns before seeing if the optimised method could be repeated on modern versions of 1985 columns. Overall, I am not persuaded that the ordinary skilled person would have achieved separation by preparative HPLC in 1985.

I must also consider what the skilled person would have done if presented with the results of the experiments. As I have mentioned, the products of all the experiments were subjected to NMR, chiral HPLC and polarimetry analyses. It is common ground that NMR spectra do not describe absolute stereochemistry and so would not provide the skilled person with any detail as to the enantiomeric purity of the samples analysed. Moreover, the chiral HPLC analyses were carried out using equipment which was not available in 1985. Nor am I satisfied that chiral HPLC was generally available at that time. So, of all the experiments carried out, the skilled person would have had only the polarimetry data to guide him. As to these, Professor Davies thought the variability of the specific rotations was such that the skilled person would have concluded the experiments were not reproducible and were therefore unreliable. Dr Spargo considered the technique itself was unreliable and ‘disowned’ the data, although it is to be noted that he did suggest the use of the technique in his protocol. In his evidence he said he would have taken each product, treated it with menthyl chloroformate to reform the diastereomer and passed it down an analytical HPLC column to assess its enantiomeric purity. Unfortunately this was not done. Dr Spargo also suggested the skilled person would have proceeded despite the lack of consistency in the data. On this point I prefer the evidence of Professor Davies. It is important to be wary of hindsight in considering these results. The chiral HPLC data reveal that the resolution was successful but, as I have found, this would not have been known to the skilled person. He would have been faced with what appeared to be another unsuccessful attempt at resolution. Would he simply have ploughed on? I do not believe he would have been minded to do so any more than with any of the earlier failed experiments. I am reinforced in this conclusion by the very low yields obtained. I think the skilled person would not have considered it worthwhile continuing with his method. If he had reached this far and was minded to continue with his attempts to resolve ofloxacin, I think he would have turned to another resolution technique.

In conclusion, it was not obvious to resolve ofloxacin and to produce levofloxacin in the light of the common general knowledge and the 005 Patent and Drugs of the Future. The enantiomers could not be separated relatively easily. To the contrary, resolution involved a research programme with a highly uncertain outcome and I consider the person skilled in the art would have turned his attention to the development of new molecules.

I now consider the attack of obviousness based upon Gerster I, Rohlfing and Gerster IP and, in each case, the common general knowledge. I will deal with them in the order in which they were addressed by the experts.

Gerster I is an abstract from the proceedings of the North American Medicinal Chemistry Symposium, Toronto, which took place in 1982. As I have mentioned, the authors were from Riker.

Gerster I explains that the Riker group designed the tricyclic structure of flumequine to study how bacterial activity would be affected by restricting the free rotation of the N-ethyl group present in many of the quinolones under development at the time. They noted that an asymmetric centre was created and explained that they had succeeded in preparing both optical isomers. They found that one isomer had potent antibacterial activity and the other only very weak activity. They also prepared the simpler desmethyl achiral analogue and showed that it had less antibacterial activity than flumequine but was more active than the weakly active isomer.

The obviousness case based upon Gerster I was explained very concisely by Dr Spargo. He considered the stereochemical structure activity observation made by Riker to be significant from a medicinal chemistry perspective, and something that the skilled person would definitely have been interested in. He would have been encouraged by Gerster I to prepare the enantiomers of chiral quinolones which are structurally related to flumequine in the expectation that they might well have different antibacterial activity.

Dr Spargo continued that ofloxacin is one such structurally related chiral quinolone, as shown in these illustrations included in his first report:

In particular, he noted that ofloxacin and flumequine share a fused tricyclic fluoroquinolonecarboxylic acid core structure although he fairly observed that flumequine has a CH₂ group instead of an oxygen atom in the saturated ring, and lacks the N-methylpiperazine substituent. Further, the chiral centres of flumequine and ofloxacin occupy the same position with respect to the tricyclic core structure.

Dr Spargo’s reasoning proceeded that, to the skilled person reading Gerster I, it would have been obvious to seek to prepare the enantiomers of ofloxacin to see whether their antibacterial activities differed significantly, as had been observed with the enantiomers of flumequine. He would have been able to locate a synthetic route to ofloxacin, as described in the 005 Patent and Drugs of the Future and would then have resolved and tested its enantiomers.

Under cross examination, Dr Spargo maintained his opinion. He emphasised this publication established that chirality does make a difference in the quinolone structure. In this sense he thought it a landmark disclosure.

Professor Wentland took the contrary position and, importantly, put flumequine in context in 1985, the relevant date for considering the question of obviousness. It must be remembered that flumequine had been known since 1973 and was recognised as an important advance as of that time. However, by 1985, its properties were considered unexceptional and it had made only limited progress as a human therapeutic. Consequently it was no longer an influential compound. I consider this is a matter I must take into account in determining what, if anything, it would have been obvious to the skilled person to do in the light of the disclosure.

He also explained that the skilled person would recognise the differences between the molecules as significant. Specifically, the additional oxygen can act as a hydrogen bond acceptor to a complementary site on the DNA gyrase cleavable complex, whereas the CH₂ group in flumequine is not capable of doing this. Moreover, the oxygen is more hydrophilic than CH₂ and this will change the electronics of the molecule. In addition, the large steric bulk of the methyl piperazine group could also make a difference to its binding ability, as could its positive charge. He maintained under cross examination that these differences were such that one could not make a deduction from one to the other. Dr Spargo agreed to an extent in that he accepted that with these differences one could not say that flumequine and ofloxacin interacted with the gyrase in a similar fashion. Finally, I should reiterate that Professor Wentland was also firmly of the view that those in the quinolone field were simply not interested in enantiomers in the first place.

Professor Davies shared Professor Wentland’s opinion that flumequine and ofloxacin were different compounds and that the medicinal chemist could not take information from one to the other. He focussed on the additional oxygen and the presence of the two additional nitrogen atoms in the piperazine group. He believed the differences were such that one could not make any prediction about the biological activity of the enantiomers and further, that these differences were likely to have a significant effect of the ADME properties of the compounds. He too maintained his position under extensive cross examination.

Weighing this evidence and taking care to avoid the dangers of hindsight as I must, I prefer the evidence of Professor Wentland and Professor Davis. Consequently I do not consider Gerster I would have rendered it obvious to the skilled person to take any step it would not already have been obvious for him to take in the light of his knowledge of ofloxacin. As I have found, by 1985 he would have been aware that ofloxacin had a chiral centre and comprised two enantiomers. He would also have appreciated that the activity of the two enantiomers might be different and that theoretically all of the activity might lie in one rather than the other. It would also have been apparent to him that there are significant structural differences between flumequine and ofloxacin and that their overall activities are very different, with ofloxacin being a markedly superior molecule. I conclude that these structural differences and the lack of any clear understanding of how the molecules bind to the gyrase target would have rendered it impossible to make a prediction as to the behaviour of the ofloxacin enantiomers on the basis of Gerster I, and the disclosure would not have rendered it obvious to take any step in relation to ofloxacin. Moreover, Gerster I provides no assistance as to how the enantiomers of ofloxacin might be resolved. Accordingly, the allegation of obviousness over Gerster I fails just as it did over the Daiichi publications.

Rohlfing was published in 1985 and was again authored by workers at Riker. It describes the effect of the various quinolones on different strains of the bacterium Neisseria gonorrhoeae. Importantly for present purposes, it includes this sentence:

"In addition to fluorination and the influence of adjacent substituents just described, the importance of configuration at position 5 (flumequine) has been established (Gerster, Rohlfing & Winandy 1982). This may be a feature of the improved potency we observed for ofloxacin and ciprofloxacin relative to their congeners."

This gave rise to a debate. Professor Wentland expressed the view the authors were reasoning that the common structural feature responsible for the improved activity of ofloxacin and ciprofloxacin is the carbon atom at position 1 (using the bicyclic numbering system) which has two additional carbons attached to it. Since ciprofloxacin, an achiral molecule, is included in the analysis with ofloxacin, a racemate, a skilled medicinal chemist would not conclude the article addresses the chirality of quinolones at all. Dr Spargo considered the reference to configuration is plainly a reference back to its use in Gerster I where it is being used to refer to whether the methyl group is in one orientation or the other.

For my part, I believe the teaching of Rohlfing is ambiguous. However, there is a clear reference back to Gerster I and the skilled person must be assumed to have read it together with Rohlfing. If he did so then he would undoubtedly have become aware of the difference in the activity between the enantiomers of flumequine. Moreover, Rohlfing clearly draws attention to the same chiral centre in ofloxacin. In the circumstances, it seems to me the ambiguity is academic and I will assume that Dr Spargo’s opinion as to how Rohlfing would have been read is correct. On that basis GUK submits that even if the skilled person was not interested in the enantiomers of ofloxacin having read Gerster I, he certainly would have been having read Rohlfing.

In my judgment, and in the light of my findings, the position in relation to Rohlfing is no better than it is in relation to Gerster I. In 1985, the skilled person would have been aware that ofloxacin had a chiral centre and comprised two enantiomers. He would also have appreciated that the activity of the two enantiomers might be different and that theoretically all of the activity might lie in one rather than the other. But Rohlfing would have provided him with no further guidance as to their particular behaviour. Nor does it provide any assistance as to how the enantiomers of ofloxacin might be resolved. The allegation of obviousness over Rohlfing fails, just as it did over the Daiichi publications.

Gerster IP was introduced into the proceedings relatively late but it is undoubtedly an important citation. It is a copy of the poster presentation corresponding to the Gerster I abstract.

Gerster IP reports the synthesis of the enantiomers and the achiral desmethyl analogue of flumequine. The authors report that one isomer had potent antibacterial activity and the other one only very weak activity. The desmethyl analogue showed less activity than flumequine but was more active than the weakly active isomer. The authors conclude that the addition of a methyl group in the right configuration enhances biological activity while a methyl group in the wrong configuration diminishes that activity. At the end of the poster the authors explain they achieved separation by forming diastereomeric amides from the racemic intermediate 6FTHQ and N-tosyl-L-prolyl chloride which is, in Dr Spargo’s opinion, a relatively unusual choice of derivatising agent. One of the resulting diastereomers was isolated by recrystallisation, while the other was isolated using preparative HPLC. Saponification of the purified diastereomeric amides gave the enantiomers of 6FTHQ, which were then used to prepare the enantiomers of flumequine.

The first point I have to consider in relation to Gerster IP is whether it was made available to the public at all and, if so, how it should be treated. The evidence before me establishes that the poster was put up by Dr Gerster at the Toronto meeting on 23 June 1982 for two hours in a public place where those passing by could see it. He was present for the whole time, standing by the poster, and he answered any questions posed to him. Dr Gerster recalled that people did come by, but I have no evidence that anyone took any interest in the poster. Dr Gerster cannot remember anyone ever requesting a copy but said he would have sent one to anybody that asked for it.

Daiichi submits that in these circumstances Gerster IP was available, if at all, for a short period in June 1982. But it was not catalogued or indexed or in any other way made capable of being retrieved by the skilled person at the priority date in 1985, by which time the common general knowledge of those in the quinolone field had changed considerably. Thus the combination of information upon which GUK seeks to rely, that is to say the teaching of Gerster IP and the common general knowledge about ofloxacin, never existed and, indeed, never could have existed. Whilst patent law can acknowledge certain legal fictions, there is no justification under the statute to support combinations of disclosures which could not, in fact, have been made.

In support of this submission, I was referred to the decision of the Court of Appeal in Windsurfing International Inc v Tabur Marine (Great Britain) Ltd 1985] RPC 59 and, in particular, to the following passage in the judgment of Oliver LJ at page 78 where, in considering whether the use of a sailboard in 1958 (many years before the priority date) made the invention obvious, he said:

“If that be right, then one asks "upon what ground could it be said that the substitution of a wishbone boom was not obvious?" Only, we think, upon the ground that the hypothetical skilled man in 1958 would have been so uninterested in this child's plaything that he would not have applied his mind to the matter at all for, on the evidence, it is, we think, clear that any skilled adult who applied his mind to Chilvers' device would at once have seen it as obvious that the unconventional and primitive split boom devised by Chilvers ought to be replaced by the conventional wishbone boom which, even though not in everyday use, would then have been familiar to anyone skilled in yacht building.”

In this passage, Daiichi says, it is plain Oliver LJ was considering the impact of the disclosure as of its date and not later. Accordingly, it is not permissible to combine the teaching of Gerster IP with the knowledge of ofloxacin that those in the field had only acquired by 1985.

In assessing this submission, I think the starting point is to consider whether Gerster IP was made available at all. In my judgment it plainly was. The poster could be read by anyone at the conference free from any obligation of confidence. In the case of information contained in a document, it is the theoretical possibility of having access to the information which makes it available and part of the state of the art. It is irrelevant as a matter of law whether a member of the public actually reads it. Thus in Lux v Pike [1993] RPC 107 Aldous J said at 133:

“…it is settled law that there is no need to prove that anybody actually saw the disclosure provided the relevant disclosure was in public. Thus an anticipating description in a book will invalidate a patent if the book is on a shelf of a library open to the public, whether or not anybody read the book and whether or not it was situated in a dark and dusty corner of the library. If the book is available to the public, then the public have the right to make and use the information in the book without hindrance from a monopoly granted by the State.”

The next question is whether information can cease to be part of the state of the art at some point after it has been made available. I do not believe that it can. Under Article 54(2) of the EPC the state of the art comprises everything made available to the public by means of written or oral description, by use, or in any other way, before the priority date. The EPC does not contemplate the possibility that once information has been made available it can subsequently be withdrawn or that it can cease to be part of the state of the art after time. Moreover, a patent monopoly cannot prevent anyone doing anything which is obvious. This must be assessed at the priority date and through the eyes of the skilled but non inventive person, a legal creation, who is deemed to be aware of the common general knowledge and is entitled to do anything which is obvious over anything made available before that date. I therefore reject Daiichi’s submission. In my judgment it is permissible to rely upon Gerster IP in combination with the common general knowledge in 1985 and consider whether it rendered the invention of the Patent obvious. Finally, I must deal with the Windsurfing case. I accept that the Court of Appeal appears to have considered the issue of obviousness over the Chilvers device some time before the priority date. But the particular point taken before me by Daiichi does not appear to have been argued in that case and there is nothing in the reasoning of the court to suggest it would have come to a conclusion contrary to my own.

Turning to the question of obviousness over Gerster IP in 1985, my reasoning in relation to Gerster I applies equally here, save that I must also consider the disclosure in Gerster IP of a method of resolution. This is important because if this method is applied to ofloxacin, it works, as process C of the Patent shows.

To summarise my conclusions thus far: flumequine was not an influential compound by 1985 and I do not believe that information about its enantiomers disclosed in a poster from 1982 would have been of any real interest to those seeking to develop an improved quinolone some three years later. Further, the significant structural differences and the lack of any clear understanding of how the molecules bind to the gyrase target would have rendered it impossible to make a prediction as to the behaviour of the enantiomers of ofloxacin. Against this background, I must consider whether it would have been obvious to the skilled person to apply the method of resolution described in Gerster IP to ofloxacin. I think the following further matters are relevant. First, on reading the method the skilled person would have seen that an unusual derivatising agent was used, namely N-tosyl-L-prolyl chloride. Second, he would have seen it was reacted with an intermediate, which, on the evidence, is not the first approach the skilled person would take to resolution, which is to seek to resolve the final racemate by making diastereomeric salts. Third, GUK’s case depends upon the skilled person recognising the similarities between 6FTHQ and intermediate (V) in Drugs of the Future. Whilst I accept that the skilled person could find out the method of making ofloxacin in 1985 without undue effort, I am not satisfied that he would have known that method as part of his common general knowledge. Accordingly, it is not established on the evidence how the skilled person would ever have come to compare 6FTHQ and intermediate (V). Fourth, even if the skilled person did come to compare the method of resolving flumequine with the method of making ofloxacin and noticed the similarity between 6FTHQ and intermediate (V), he would also have noticed the differences. As Professor Davies explained, both are anilines which means that the nitrogen to which the resolving agent is to be attached is a poor nucleophile and the position is made worse in the case of intermediate (V) by the presence of the additional oxygen and fluorine atoms. As he maintained under cross examination, it might or might not work and the skilled person would be faced with a huge choice of reactions to try. Moreover, there were other reactions staring the skilled worker in the face, namely those involving diastereomeric salts. Dr Spargo was much more enthusiastic. He thought the skilled person would regard Gerster IP as a strong encouragement, would have every reason to think N-tosyl-L-prolyl chloride would work with intermediate (V) and would give it a try as a reasonable reaction. I should also mention that GUK relied on the fact that Dr Hayakawa apparently made the invention after seeing Gerster II, which provides similar teaching in relation to the flumequine analogue S-25930 which differs in the presence of an extra methyl group.

This last point seems to me to add little weight, Dr Hayakawa being one of the inventors. However, I have given particularly careful consideration to the opinions of Dr Spargo, which he gave clearly and fairly. But I am conscious too of the need to avoid hindsight. It is all too easy now the invention has been described to see how it could have been made by applying the teaching of Gerster IP, using N-tosyl-L-prolyl chloride to make covalent diastereomers of intermediate (V) and then resolving them. Standing back and considering all the matters to which I have referred in the immediately preceding paragraph against the background of the common general knowledge and the extent of the interest I believe the skilled person would have had in trying to separate the enantiomers of ofloxacin, I have reached the conclusion that it was not obvious to make levofloxacin in the light of Gerster IP.

Gerster II only becomes relevant if the Patent is not entitled to its priority date. Moreover the same arguments apply as for Gerster IP. No separate arguments were advanced in relation to it and I need say no more about it.

Daiichi relies on the commercial and technical success of levofloxacin since it was launched. It also says, if levofloxacin was obvious, and bearing in mind ofloxacin was disclosed in the 005 Patent in 1983, why was it not made before? There can be no doubt the commercial success of levofloxacin has been substantial. In 2007 it achieved sales in the US of US $1.76 billion and in the EU of in excess of US $290 million. Moreover, as Professor Zhanel explained, it has broad spectrum antimicrobial activity, maintains the Gram-negative and improves upon the Gram-positive activity of ciprofloxacin and can be safely administered in relatively high doses of 1,000 mg.

Impressive though these matters appear, in the circumstances of this case I prefer not to place weight on them for the following reasons. So far as the sales figures are concerned, I have no evidence of the context in which they were achieved. GUK fairly asks, how do the figures compare with those of other anti-bacterials? How much effort has been put into switching patients from ciprofloxacin to levofloxacin? What were the relative prices of ciprofloxacin and levofloxacin? Indeed, relative to marketing spend, it seems that sales of ofloxacin have been better than those of levofloxacin. Turning to the technical success of levofloxacin, had I concluded it was obvious to make levofloxacin then I do not believe it would have ceased to be obvious merely because it turned out to be better than expected. Moreover, it may well be that part of its technical and commercial success is attributable to the features that Daiichi had not discovered at the first or second priority dates, namely its toxicological profile and solubility characteristics, and to which, for the reasons I have given, I do not consider it is permissible to attach weight.

GUK contends that if any technical contribution has been made by Daiichi then it was in finding a way to obtain a single enantiomer of ofloxacin, which was a known desideratum. Claim 2 extends beyond that technical contribution and covers all ways of achieving the known desideratum. Accordingly, it is insufficient. Further, having regard to the fact that the 005 Patent and Drugs of the Future disclose a method of making ofloxacin that comprises as a final step reacting the racemic version of compound (V) with an N-alkyl piperazine of the formula shown in claim 5 of the Patent, obtaining compound (V) was itself also a known desideratum as a resolved final stage intermediate for the purpose of making levofloxacin. Claim 5, it is said, extends beyond any technical contribution made by the alleged invention and covers all ways of making levofloxacin which have as a step the reaction of compound (V) with an N-alkyl piperazine of the formula shown in claim 5. Accordingly, it too is insufficient.

As I have indicated, GUK accepts this attack must fail in this court in the light of the decision of the Court of Appeal in Lundbeck. Nevertheless, I am invited to make findings of fact in case the decision in Lundbeck should be reversed on further appeal. As for claim 2, I have found that the skilled person would have considered it worthwhile exploring whether the enantiomers of ofloxacin could be separated relatively easily. By that I have in mind the application of routine techniques to ofloxacin. I do not consider it was obvious to pursue this investigation in the manner of the experiments relied upon by GUK; nor am I satisfied it was obvious to investigate the resolution of compound (V).

The point goes only to claim 5 and it is this. The application as filed did not contain claim 5 or any part of the description corresponding to claim 5. It contained merely processes A, B and C. GUK accepts that each of these processes embodies a common feature, namely that the last step involves adding N-methyl piperazine to the final intermediate product. But GUK says the application did not suggest that step had any inventive significance. In the patent as granted processes A, B and C have been generalised out to the generic process of claim 5 using that step as a distinguishing and unifying feature when it was not suggested as having any such significance in the application.

In Bonzel v Intervention (No 3) [1991] RPC 553 Aldous J stated the test for examining whether an amendment involves the adding of subject matter at 574:

“(1)

To ascertain through the eyes of the skilled addressee what is disclosed, both explicitly and implicitly in the application.

(2)

To do the same in respect of the patent as granted.

(3)

To compare the two disclosures and decide whether any subject matter relevant to the invention has been added whether by deletion or addition. The comparison is strict in the sense that subject matter will be added unless such matter is clearly and unambiguously disclosed in the application either explicitly or implicitly.”

In Richardson-Vicks' Patent [1995] RPC 568 at 576 Jacob J summarised the rule this way:

“I think the test of added matter is whether a skilled man would, upon looking at the amended specification, learn anything about the invention which he could not learn from the unamended specification.”

In Vector v Glatt [2007] EWCA Civ 805 Jacob LJ elaborated the position in relation to intermediate generalisations:

“A particular, and sometimes subtle, form of extended subject matter (what our Act calls “additional matter”) is what goes by the jargon term “intermediate generalisation”. Pumfrey J described this in Palmaz's European Patents [1999] RPC 47 , 71 as follows:

“If the specification discloses distinct sub-classes of the overall inventive concept, then it should be possible to amend down to one or other of those sub-classes, whether or not they are presented as inventively distinct in the specification before amendment. The difficulty comes when it is sought to take features which are only disclosed in a particular context and which are not disclosed as having any inventive significance and introduce them into the claim deprived of that context. This is a process sometimes called “intermediate generalisation.””

Claim 5 is directed to a way of making only one of two compounds, levofloxacin or its ethyl analogue. The method involves reacting intermediate (V) as drawn (in which X₂ represents a halide) with an N-alkyl piperazine. GUK makes no complaint of the reference to X₂ as a halide or N-alkyl piperazine as such, but says that Daiichi has drawn a single aspect out of the processes A, B and C which was not suggested to have significance as possessing a distinguishing and unifying quality. It has therefore claimed a different invention from that disclosed in the application and generalised a single feature in an impermissible way.

I am unable to accept this submission. The process of claim 5 forms a part of each of processes A, B and C and it would be immediately apparent to the skilled person that it could be applied to intermediate (V) however that intermediate has been made. This is not a case where a feature has been disclosed only in a particular context and then introduced into a claim deprived of that context. Nor does it matter that it was not presented in the application as being inventively distinct. The application disclosed a method of converting intermediate (V) to levofloxacin or its ethyl analogue. The skilled person would immediately recognise it to be of general application and in my judgment the inclusion of a claim to that method does not involve the addition of subject matter.

I must now consider the final attack on the SPC. GUK contends the UK marketing authorisations relied upon in support of the application for the SPC were not the first authorisations to place the product on the market as a medicinal product.

The relevant background facts are these:

The SPC Regulation arose from a recognition of the need to provide an effective period of protection for research centres developing medicinal products, as is readily apparent from the following recitals (with numbering added):

“(1)

Whereas pharmaceutical research plays a decisive role in the continuing improvement in public health;

(2)

Whereas medicinal products, especially those that are the result of long, costly research will not continue to be developed in the Community and in Europe unless they are covered by favourable rules that provide for sufficient protection to encourage such research;

(3)

Whereas at the moment the period that elapses between the filing of an application for a patent for a new medicinal product and authorization to place the medicinal product on the market makes the period of effective protection under the patent insufficient to cover the investment put into the research;

(4)

Whereas this situation leads to a lack of protection which penalizes pharmaceutical research;

(5)

Whereas the current situation is creating the risk of research centres situated in the Member States relocating to countries that already offer greater protection;”

The solution was the creation of a scheme of supplementary protection certificates granted, under the same conditions, by each Member State at the request of the holder of a national or European patent relating to a medicinal product for which a marketing authorisation has been obtained. The intended duration of protection is explained in recitals (8) and (9):

“(8)

Whereas the duration of the protection granted by the certificate should be such as to provide adequate effective protection; whereas, for this purpose, the holder of both a patent and a certificate should be able to enjoy an overall maximum of fifteen years of exclusivity from the time the medicinal product in question first obtains authorization to be placed on the market in the Community;

(9)

Whereas all the interests at stake, including those of public health, in a sector as complex and sensitive as the pharmaceutical sector must nevertheless be taken into account; whereas, for this purpose, the certificate cannot be granted for a period exceeding five years; whereas the protection granted should furthermore be strictly confined to the product which obtained authorization to be placed on the market as a medicinal product;”

Against this background, I come to the relevant Articles. The definitions are to be found in Article 1:

Article 1

Definitions

For the purposes of this Regulation:

(a)

'medicinal product' means any substance or combination of substances presented for treating or preventing disease in human beings or animals and any substance or combination of substances which may be administered to human beings or animals with a view to making a medical diagnosis or to restoring, correcting or modifying physiological functions in humans or in animals;

(b)

'product' means the active ingredient or combination of active ingredients of a medicinal product;

(c)

'basic patent' means a patent which protects a product as defined in (b) as such, a process to obtain a product or an application of a product, and which is designated by its holder for the purpose of the procedure for grant of a certificate;

(d)

'certificate' means the supplementary protection certificate.”

The conditions for obtaining a certificate are set out in Article 3:

Article 3

Conditions for obtaining a certificate

A certificate shall be granted if, in the Member State in which the application referred to in Article 7 is submitted and at the date of that application:

(a)

the product is protected by a basic patent in force;

(b)

a valid authorization to place the product on the market as a medicinal product has been granted in accordance with Directive 65/65/EEC or Directive 81/851/EEC, as appropriate;

(c)

the product has not already been the subject of a certificate;

(d)

the authorization referred to in (b) is the first authorization to place the product on the market as a medicinal product.”

The subject matter of protection and rights conferred are addressed by Articles 4 and 5:

Article 4

Subject-matter of protection

Within the limits of the protection conferred by the basic patent, the protection conferred by a certificate shall extend only to the product covered by the authorization to place the corresponding medicinal product on the market and for any use of the product as a medicinal product that has been authorized before the expiry of the certificate.

Article 5

Effects of the certificate

Subject to the provisions of Article 4, the certificate shall confer the same rights as conferred by the basic patent and shall be subject to the same limitations and the same obligations.

Finally, the duration of protection is specified in Article 13:

Article 13

Duration of the certificate

1.

The certificate shall take effect at the end of the lawful term of the basic patent for a period equal to the period which elapsed between the date on which the application for a basic patent was lodged and the date of the first authorization to place the product on the market in the Community reduced by a period of five years.

2.

Notwithstanding paragraph 1, the duration of the certificate may not exceed five years from the date on which it takes effect.

It will be seen that the scheme of the SPC Regulation reflects the recitals. Protection is to be afforded to a product which is protected by a basic patent, has a marketing authorisation as a medicinal product and has not already been the subject of an authorisation or certificate. Protection is conferred for a maximum period of 15 years from the date of the first authorisation, including the life of the basic patent.

In this case the dispute primarily concerns the conditions specified by Article 3. GUK accepts that conditions (a)-(c) were satisfied but contends (d) was not. It says levofloxacin is clearly a “product”. But if one asks: “Were the 6 June 1997 authorisations the first authorisations to place levofloxacin on the market in the UK as a medicinal product?” it says the answer is no. There was a previous marketing authorisation to place it on the market, dated 16 March 1990. In that case it was authorised for marketing only in the form of a 50:50 mixture together with the (+) enantiomer of ofloxacin, but it was still an authorisation to put levofloxacin on the market as a medicinal product.

GUK also puts its case another way on the basis of the German authorisation. It says Article 13 sets the term for an SPC by reference to the date of the first authorisation to place the product on the market in the Community. In the present case, there was a German marketing authorisation for ofloxacin granted on 31 May 1985. If GUK is right about the meaning of “product” then the term of the SPC according to Article 13 should have been zero.

My immediate reaction is that these contentions are misconceived. It was not until June 1997 that any marketing authorisation was granted to place the product levofloxacin on the market as a medicinal product. The authorisations of May 1985 and March 1990 were not granted in respect of levofloxacin, but rather in respect of a different product, namely ofloxacin. Nevertheless, I have been taken to a number of decisions of the European Court of Justice (the “ECJ”) and of this court which are said to compel me to arrive at the opposite conclusion.

In Case C-392/97 Farmitalia Carlo Erba SRL’s Supplementary Protection Certificate Application [2000] RPC 580, Farmitalia secured patent protection for a drug called idarubicin and subsequently obtained marketing authorisations for two medicinal products for the treatment of particular leukaemias in humans, in which the active ingredient was idarubicin hydrochloride and the ancillary ingredient was dehydrated lactose. It thereupon sought a certificate for “idarubicin and salt thereof including idarubicin hydrochloride”. However, the German patent office was only prepared to grant a certificate for the medicament containing the particular salt idarubicin hydrochloride. The decision was upheld by the Bundespatentgericht but on further appeal the Bundesgerichtshof referred the issue to the ECJ for a preliminary ruling on the proper interpretation of Article 3(b). The question asked was, in substance, whether, on a proper interpretation of Article 3(b), the certificate can protect the product only in the specific form mentioned in the marketing authorisation.

Farmitalia argued that the basic patent protected the active ingredient and its salts, whereas the certificate granted by the German patent office protected only the particular salt form of the active ingredient mentioned as the active ingredient in the marketing authorisation. This would allow competitors, after the basic patent had expired, to obtain marketing authorisations for a different salt of the same active ingredient and so compete with therapeutically equivalent products. The Court accepted the argument and considered that if the certificate did not cover the actual medicinal product, as protected by the basic patent and one of the possible forms of which was the subject matter of a marketing authorisation, the fundamental objective of the SPC Regulation, as set out in the first and second recitals, could not be attained.

Accordingly the decision establishes that, where a product in the form referred to in the marketing authorisation is protected by a basic patent in force, the certificate is capable of covering that product, as a medicinal product, in any of the forms enjoying the protection of the basic patent. This, however, is a rather different point to that which I have to decide. There is no dispute that Daiichi could have obtained an SPC for any form of ofloxacin founded upon the 005 Patent.

In Case C-258/99 BASF AG v Bureau Voor De Industriele Eigendom [2002] RPC 9 the ECJ had to consider whether products differing only in their impurity levels were different products within the meaning of the Plant Protection Regulation 1610/96 (“the Plant Regulation”). In 1967, BASF was granted an authorisation for the pesticide “Pyramin, containing chloridazon as an active substance”. Chloridazon in fact comprised two isomers, one of which was active and the other of which was inactive and regarded as an impurity. In 1982, BASF was granted a patent for a process for making chloridazon which enabled a greater proportion of active to inactive isomer to be obtained. In 1987, it was then granted another authorisation for the pesticide “Pyramin DF, containing chloridazon as an active substance” with an improved ratio of active to inactive isomers reflecting the new patented process. In 1997, BASF applied to the Dutch patent office for a supplementary protection certificate for chloridazon as a plant protection product. It was refused. On appeal, a number of questions were referred to the ECJ for a preliminary ruling. Of these, the first question is the most important for present purposes and it was broken down into three parts.

By the first part of the first question, the national court essentially sought a definition of the concept of a product within the meaning of the Plant Regulation. The Court answered that the concept of a product within the meaning of the Plant Regulation covers chemical elements and their compounds, as they occur naturally or by manufacture, including any impurity inevitably resulting from the manufacturing process, which have general or specific action against harmful organisms or on plants. I should say that in giving this answer, the Court drew on the definition of product as an active substance, and that ‘substance’ is itself defined in the Plant Regulation in a way which has no direct counterpart in the SPC Regulation. Nevertheless, I do not consider this to be a material distinction.

By the second part of the first question the national court essentially asked whether two products which differ only in the proportion of the active chemical compound to the impurity they contain, must be regarded as the same product. Not surprisingly, the Court answered that they must.

By the third and final part of the first question the court was asked whether the fact that a new marketing authorisation must be obtained for the new plant protection product which has a different proportion of active chemical compound to impurity from that of the former plant protection product is relevant for the purposes of establishing whether or not the constituent products of those plant protection products are the same. The Court answered that it is not. The marketing authorisation is not among the criteria used by the Plant Regulation for defining the concept of a product.

It clearly emerges from this decision that the focus of the enquiry must be on the product. As has been seen, in the SPC Regulation, this is defined as the active ingredient or combination of active ingredients. Two medicinal products may have the same active ingredient, and hence comprise the same product, even though they have different quantities of impurities and so require separate marketing authorisations. This has a direct bearing on the issue before me and I must return to consider its implications after addressing the other cases to which I was referred.

I was then taken to Case C-431/04 Massachusetts Institute of Technology [2006] FSR 34. MIT was the holder of a patent covering the combination of two elements, polifeprosan, a polymeric, biodegradable excipient, and carmustine, an active ingredient already used in chemotherapy for the treatment of brain tumours. The commercial embodiment of the invention was called Gliadel, a device implanted into the cranium where it slowly released carmustine upon bioerosion of the polifeprosan matrix.

A marketing authorisation for Gliadel was granted in Germany and, in reliance upon that authorisation, MIT asked the German patent office to grant it an SPC. In its main application it requested that an SPC be granted for carmustine in combination with polifeprosan. Its alternative application sought an SPC for carmustine alone. The German office rejected the main application on the basis that polifeprosan could not be considered to be an active ingredient. It also held that no SPC could be granted for carmustine alone because that active ingredient was already covered by an SPC, and had been for a long time. MIT appealed to the Bundespatentgericht and thence to the Bundesgerichtshof which stayed the proceedings and referred a question to the ECJ for a preliminary ruling. It essentially asked whether Article 1(b) of the SPC Regulation must be interpreted so as to include in the concept of “combination of active ingredients of a medicinal product” a combination of two substances, only one of which has therapeutic effects of its own for a specific indication, the other rendering possible a pharmaceutical form of the medicinal product which is necessary for the therapeutic efficacy of the first substance for that indication. The Court noted that ‘active ingredient’ is generally accepted in pharmacology not to include substances which do not have an effect of their own on the human or animal body, such as excipients. It therefore ruled the alliance of such a substance with a substance which does have therapeutic effects of its own cannot give rise to ‘a combination of active ingredients’ within the meaning of the SPC Regulation.

So the MIT case establishes that an ingredient or a combination of ingredients cannot constitute a product within the meaning of the SPC Regulation unless it has (or they have) activity, that is to say an effect on the human or animal body.

The last decision of the ECJ to which I was referred was Case-202/05 Yissum Research and Development Company of the Hebrew University of Jerusalem v Comptroller General of Patents (17 April 2007). This establishes a short point: the concept of ‘product’ does not include the therapeutic use of an active ingredient. It is not one which arises in the case before me.

I was also taken to Draco’s Application [1996] RPC 417, a decision of Jacob J. In 1973, Draco applied for and was granted a patent which covered budenoside, a compound used in the treatment of asthma. In 1981, a product authorisation was obtained for a particular aerosol formulation and in 1982, a further product authorisation was obtained for essentially the same formulation but in a different dose. After further work, Draco developed a method of delivery using a device called a turbuhaler which did not involve an aerosol but instead allowed budenoside to be administered in the form of micronised particles. This required a third product authorisation which was obtained in 1990. It thereupon sought an SPC which was refused. As Jacob J explained, the scheme of the SPC Regulation is to provide a period of exclusivity under the patent and the SPC of a maximum of 15 years from the date of the first authorisation of the product, yet the effect of the SPC sought would have been to cover not only the precise presentation the subject of the third product authorisations, but also those permitted by the first and second authorisations too. So the period of protection would have been well over 15 years.

In the light of these cases, I believe the fundamental question is whether or not the product the subject of the 1985 and 1990 authorisations was ofloxacin or levofloxacin.

In my judgment the product the subject of the 1985 and 1990 authorisations was indeed ofloxacin and the authorisation of 1997 was the first authorisation to place levofloxacin on the market as a medicinal product. I reach that conclusion for the following reasons.

First, ofloxacin itself was recognised to be a good antimicrobial agent. It had been studied and had been shown to be active against a broad spectrum of microbes, maintaining potent Gram-negative activity with improved activity against Gram-positive organisms compared to norfloxacin. Moreover, its solubility and toxicity characteristics, amongst others, had been determined. It was known to be a chiral molecule and hence to comprise two enantiomers but, until the invention of the Patent, it was not known what the characteristics of the enantiomers were.

Second, ofloxacin was recognised to be the active ingredient of the medicinal products for which authorisations were sought and duly granted in 1985 and 1990. Those authorisations constituted permission to place ofloxacin, not levofloxacin, on the market as a medicinal product and the active ingredient was duly named as ofloxacin.

Third, it is now known that ofloxacin, levofloxacin and the R(+) enantiomer each have different properties. As I have explained, it is taught in the Patent that levofloxacin has an antimicrobial activity of about twice that of ofloxacin, markedly greater solubility and less toxicity. The R(+) enantiomer has an antimicrobial activity of about 1/10 to 1/100 that of ofloxacin but has markedly greater solubility and approximately the same toxicity. Further, these characteristics of the R(+) enantiomer must be seen in context. As Professor Zhanel pointed out, it is still more active than nalidixic acid and pipemidic acid. Indeed, it is also almost as active as flumequine against Gram-negative organisms. It has also been hypothesised that there may be enantiomer-enantiomer interactions when the racemate is administered.

I think it follows from the foregoing that an authorisation to place ofloxacin on the market as a medicinal product cannot be considered an authorisation to place levofloxacin on the market as a medicinal product. Ofloxacin is itself an active ingredient or at least a combination of two active ingredients, namely levofloxacin and the R(+) enantiomer. All three compounds have different therapeutic effects. The R(+) enantiomer is not an inactive impurity as in the BASF case; nor is it an excipient as in the MIT case.

Moreover, it seems to me that this conclusion is consistent with the scheme and objects of the SPC Regulation. In my judgment, it required invention to make levofloxacin. The application for the Patent was filed in 1986 but authorisation to place levofloxacin the market as medicinal product was not granted until some 11 years later, in 1997. If this action is successful Daiichi will have had only nine years of protection from the date of the authorisation to the date of expiry of the Patent, and this is precisely the vice at which the SPC Regulation is aimed, as recitals (1) to (5) make clear. Furthermore, the SPC does not result in an undue extension of protection contrary to the intention of the Council expressed in recitals (8) and (9) because the SPC does not give protection against the sale of ofloxacin, a product which is accepted by GUK not to fall within the scope of the Patent.

The claim for a declaration of invalidity of (or rectification of the register in respect of) the SPC and for a declaration that claims 1-2 and 5-7 of the Patent were invalid fails. I will hear argument as to the form of order if it cannot be agreed.