[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales Court of Appeal (Civil Division) Decisions
You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> Omnipharm Ltd v Merial [2013] EWCA Civ 2 (23 January 2013) URL: http://www.bailii.org/ew/cases/EWCA/Civ/2013/2.html Cite as: [2013] EWCA Civ 2

[New search] [Printable RTF version] [Help]

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
CHANCERY DIVISION (PATENTS COURT)
THE HON MR JUSTICE FLOYD
[2011] EWHC 3393 (Pat)

		Royal Courts of Justice Strand, London, WC2A 2LL
		23/01/2013

Lord Justice Kitchin:

Introduction

1. This appeal concerns a judgment of Floyd J dated 21 December 2011 and his consequential order dated 2 March 2012 following the trial of an action brought by the respondent (Omnipharm) to revoke two patents owned by the appellant (Merial) and to secure declarations of non-infringement in respect of certain products it claimed it wished to sell.
2. The first patent, EP (UK) No. 0,881,881 (the 881 patent), relates to "spot-on" formulations of the parasiticide fipronil for protecting small animals against ticks and fleas. It has a priority date of 29 September 1995. The second patent, GB No. 2,317,564 (the 564 patent), relates to spot-on formulations for the same purpose but comprising fipronil in combination with an insect growth regulator or IGR. It has a priority date of 29 March 1996, a little later than that of the 881 patent but nothing turns on the difference.
3. The patents protect various formulations which Merial sells under the trade name Frontline. These formulations constitute a very important part of Merial's business with UK sales of at least £24 million and global sales of about $1 billion each year.
4. Omnipharm attacked both patents on the grounds of obviousness and insufficiency. Merial responded with an application, resisted by Omnipharm, to amend the patents by limiting the scope of their claims.
5. The judge held that the proposed amendments to the 881 patent were allowable and rejected the obviousness and insufficiency attacks against it. He therefore declared it valid. He also granted Omnipharm declarations of non infringement in respect of three out of four of its proposed formulations.
6. The judge also rejected the obviousness attack against the 564 patent. However, the judge found this patent was insufficient, holding that its teaching was inadequate and did not enable the skilled team to whom it is addressed to make products across the scope of its claims without undue difficulty. He therefore declared it to be invalid.
7. Upon this appeal, Merial, represented by Mr Andrew Waugh QC and Mr Thomas Hinchliffe, contends the judge fell into error in finding the 564 patent invalid for insufficiency. It recognises that such a finding is a kind of jury question which should be treated with appropriate respect by an appellate court but argues that the judge made errors of principle in reaching his conclusion which require this court to reverse his finding and reconsider the issue.
8. Merial also argues that the judge erred in principle in the way he approached the issue of costs and invites us to vary that part of the judge's final order which required Merial to pay 40% of Omnipharm's costs, excluding the costs of various interim applications. For that purpose Merial sought and we granted limited permission to appeal, as I shall explain.
9. Finally, Merial asks us to deal with one other outstanding matter concerning the costs incurred in relation to applications which it made for orders that sums paid by Omnipharm into court be maintained as security for Merial's costs pending the determination of this appeal.
10. Omnipharm, represented by Mr Henry Carr QC and Mr Tom Mitcheson, submits that the judge was right on the issue of the insufficiency of the 564 patent, essentially for the reasons he gave and for one further reason which he rejected. Omnipharm also has a contingent cross-appeal: it contends the judge's finding that the 564 patent was insufficient was consequent upon his reasons for rejecting the obviousness attack against the patent. Accordingly, so the argument goes, if the judge had not found the 564 patent insufficient he would or ought to have found it invalid for obviousness. Again, by way of contingent cross-appeal, Omnipharm contends that, by parity of reasoning, the judge ought also to have found the 881 patent invalid for obviousness.
11. At the outset of the appeals we decided to hear full argument on the issue of the insufficiency of the 564 patent. Having heard those arguments, we formed the view that the judge's decision on insufficiency must be upheld and so informed the parties, indicating that we would dismiss the appeal for reasons we would put in writing in the usual way. We also indicated it was our understanding that it was therefore not necessary to hear argument on the cross-appeal, and the parties agreed. Accordingly, we proceeded to hear further argument only on the outstanding issues concerning costs.
12. In this judgment I therefore address:

i) Merial's appeal against the judge's finding and consequential declaration that the 564 patent is invalid for insufficiency;

ii) Merial's appeal against the order requiring it to pay 40% of Omnipharm's costs; and

iii) the costs of the applications by Merial that sums paid into court by Omnipharm be maintained as security pending this appeal.

The addressee of the patents and the expert witnesses

13. There was no dispute that the patents are addressed to a notional skilled but unimaginative team working in a veterinary healthcare company interested in the development of pesticides and their formulation, and that such a team would comprise a veterinary parasitologist and a formulation scientist.
14. Each side therefore called two expert witnesses to assist the judge as to the knowledge, understanding and attitudes of this skilled team, one veterinary parasitologist and one formulation scientist.
15. Omnipharm called Mr Peter Watson as its expert in veterinary parasitology. Mr Watson had worked in a mixed animal practice before joining Bayer Animal Health where he remained until 2002. The judge rejected various attacks made by Merial on Mr Watson's evidence, finding that he did his best to put himself into the position of the ordinary skilled veterinary parasitologist without drawing upon any special knowledge internal to Bayer.
16. Omnipharm called Dr Ken Walters as its formulation expert. Dr Walters has a distinguished academic background and many years' experience in the pharmaceutical industry as a transdermal delivery scientist. As the judge held, Dr Walters is and was a leading figure in the field of the delivery of drugs through the skin and the judge had no doubt that Dr Walters genuinely held the views he expressed although, as I shall explain, the judge considered that not all of those views would have been shared by the notional skilled team.
17. Merial called Professor Michael Dryden as its expert in veterinary parasitology. He is a Distinguished Professor of Veterinary Parasitology at Kansas State University, has published widely on flea biology and, as the judge observed, is such a leading figure in this field that he has become known as "Dr Flea".
18. Finally, Merial called Professor Hadgraft as its formulation expert. He is an Emeritus Professor of Biophysical Chemistry at the School of Pharmacy at the University of London. He too has a distinguished academic background and struck the judge as being, by and large, a very rigorous scientist who sought out the evidential base for any given scientific proposition. But, as in the case of Dr Walters, there were aspects of Professor Hadgraft's evidence which the judge accepted he genuinely held but which he did not think would have been shared by the ordinary skilled team. The judge also observed that there were occasions when Professor Hadgraft's answers were rather combative and inadequately thought out and, as a result, the judge treated his evidence with a degree of caution and endeavoured to assess the extent to which it was supported by the documents.

The technical background and common general knowledge

19. The judge set out the technical background from [14]-[38] in terms with which neither side took issue. The following matters are of particular relevance to this appeal and, save where I indicate otherwise, were all accepted to form part of the common general knowledge, that is to say all that knowledge which was generally known and regarded as a good basis for further action by those in the field.

Fleas

20. Fleas are ectoparasites and they cause irritation and disease. It is therefore important to eliminate them so far as possible. Their life cycle begins with eggs and proceeds through larvae which live in dark areas under furniture or in carpets where they feed on organic debris. In time they mature into adult fleas which infect a host animal, feed on the blood of the host, mate and lay more eggs which fall to the ground, and so the cycle begins again.

Parasiticides and their application

21. Parasiticides are chemical agents which kill parasites. They fall into two broad classes, adulticides which kill the adult parasite and IGRs which interfere with the normal development of the parasite. Generally speaking, to be effective a parasiticide must be distributed, in one way or another, over the whole of the animal's body.
22. Common adulticides at the priority date included synthetic pyrethroids such as permethrin and two relatively new classes of compounds called neo-nicotinoids and pyrazoles. Fipronil, the adulticide of particular relevance in this case, is a pyrazole.
23. IGRs take longer to act than adulticides and act by breaking the life cycle of the parasite and preventing re-infection. A well known IGR at the priority date was called methoprene and it was sold in combination with a pyrethroid under the name Ovitrol Plus.
24. Parasiticides can be applied to an animal in a variety of ways. Sprays were common but suffer from the drawbacks that they require multiple actions by the user and animals do not like them very much. Nevertheless, as the judge found at [23], Merial launched fipronil as a pump action spray in France in 1994. It was called Frontline spray and rapidly became known as a very effective product, indeed better than anything else on the market.
25. Other product forms included pour-ons, which are applied by pouring the product along the midline of the back of the animal, and spot-ons, which are applied to a small area of the animal's back, usually between the shoulder blades. There were essentially two active ingredients used in spot-on formulations in 1995: fenthion, an organophosphate, and permethrin.
26. The concepts of systemic and non-systemic distribution and the differences between them are of some importance in this case and the judge dealt with them at [24]-[27]. In summary, complete coverage of the whole of an animal by a parasiticide may be achieved in a number of different ways. First, it may be applied over the whole surface of the animal by, for example, spray, shampoo or dip.
27. Second, it may be distributed systemically through the animal's blood stream, and for this purpose administered orally or transdermally, that is to say by transmission into the bloodstream though the layers of the skin. The judge termed this latter form of application and distribution "transdermal distribution". Fenthion, applied as a spot-on, was known to act in this way.
28. Third, it was generally understood by those working in the field that it is possible to apply a quantity of a parasiticide to only one part of an animal's body and find that it spreads over the whole of its body but without passing into its blood stream. Permethrin, also applied as a spot-on, was thought to act in this way.

Transdermal distribution

29. The skin of mammals has three layers, the epidermis, the dermis and the subcutaneous tissues. The dermis lies beneath the epidermis and is a vascularised tissue. So, if a drug is able to pass through the epidermis and into the dermis it can enter the animal's blood stream. An important component of the epidermis is the stratum corneum which consists of cells composed mainly of insoluble bundled keratin encased in an envelope and embedded in a lipid lamellae matrix. This plays a vital role in allowing a drug to be administered transdermally, as the judge explained at [37]-[38]:

"37. As indicated above, a compound may be delivered transdermally for systemic distribution in the host animal's bloodstream. The primary requirement for a compound to penetrate into the skin is the ability of the drug to leave the delivery vehicle and diffuse into the stratum corneum. This will depend on the partition coefficient (relative affinity) of the drug between the stratum corneum lipids and the vehicle in which the compound is carried.

38. It was known that a parameter called log P, which is the log of the partition coefficient of the compound in question, is a measure of the suitability of a drug for transdermal delivery. A log P which is less than 1 indicates a hydrophilic/lipophobic compound, whereas a log P greater than 4 would indicate a hydrophobic/lipophilic compound. Ideally, for transdermal delivery, log P should be in the intervening region, say 1-3, and preferably about 2. "

30. So it was known that a drug should have a log P of about 1-3 for transdermal distribution.

Dermal non-systemic distribution

31. As I have said, and as the judge elaborated at [43], there existed on the market in 1995 at least one spot-on formulation of a parasiticide for cats and dogs which was thought to act non-systemically. There were also spot-ons and pour-ons for larger animals which were thought to act in the same way. So it was generally understood that the application of one of these formulations to a single point on the animal's skin or along a line down the back of the animal led to its distribution over the whole of the animal's body without the formulation ever entering the animal's bloodstream. But how this distribution occurred was something of a mystery and, even today, is not fully understood.
32. Dr Walters gave evidence that the skilled team would nevertheless take steps to ensure that the formulation entered the stratum corneum and stayed there, allowing it to diffuse laterally. Professor Hadgraft took issue with this and expressed the opinion, accepted by the judge, that the skilled team would consider that passive diffusion through the stratum corneum could not account for the distribution over the whole of the animal which was seen in practice. Further, the skilled team was not equipped with any general knowledge about how to design a formulation which could be delivered through the stratum corneum in the manner suggested by Dr Walters.
33. But Professor Hadgraft went a good deal further. He expressed the view that any spot-on must act systemically and that it must enter the animal's bloodstream from the site of application and subsequently become concentrated on the surface of the skin of the animal. The judge rejected this theory too, saying it did not represent the thinking of the skilled team.
34. So although it was generally understood that some parasiticides acted non-systemically, the skilled team would not know how they worked, as the judge concluded:

"59. What I am left with, therefore, is that the skilled team would know of the existence of spot-on formulations of certain actives which were generally thought to act non-systemically. The team would not have any common general knowledge theory as to how such formulations in fact spread over the body of the animal."

35. This is an important finding and one which played a significant part in the judge's finding that the 564 patent was invalid for insufficiency.

Excipients

36. Finally I should mention that the skilled team would have had at their disposal a variety of excipients to assist in the formulation and delivery of any parasiticide formulation. These would have included various different solvents and solvent mixtures suitable for dissolving drugs with different physico-chemical characteristics and making formulations less viscous, and surfactants which could be used for various purposes including solubilising materials, enhancing their ability to penetrate a surface and enhancing their wetting and spreading properties.

The patents

37. I must now say a little about the disclosure of each of the patents and the differences between them.

The 881 patent

38. The 881 patent is entitled "Anti-parasitic composition for treating and protecting pets". The purposes of the invention are explained at [0010] – [0014] as being the provision of new compositions which are easy to use; do not require spraying of the animal's whole body; will spread over the animal's whole body following local application; and do not affect the appearance of the animal's coat, make it sticky or leave any crystals.
39. These objects are said to be met by the compositions of the invention which include fipronil at 1-20% weight per unit volume of the composition; a crystallisation inhibitor which passes a particular test called a glass plate test; an organic solvent of particular characteristics; and an organic co-solvent, again with particular characteristics.
40. At [0023] the specification explains:

"The compositions according to the invention, intended for pets, especially dogs and cats, are generally applied by depositing on the skin (in English "spot on"); it is generally a question of localised application on a surface area of less than 10 sq. cm., especially between 5 and 10 sq. cm, in particular at two points and preferably located between the animal's shoulders. After being put on, the composition spreads, particularly over the animal's whole body, and then dries, without crystallisation and without changing the appearance (in particular there is no whitish deposit or dusty appearance), or the feel of the animal's coat."

41. A long list of solvents is set out at [0025], with a shorter list of preferred solvents at [0026]. Then at [0027] there is a long list of crystallisation inhibitors. At [0028] – [0032] it is said that in a preferred embodiment a pair of crystallisation inhibitors is used, that is to say a combination of a film-creating agent of the polymeric type and a surface tension agent. Three possible co-solvents are identified at [0033].
42. The description then describes a series of 24 examples which illustrate the invention and how it can be put into practice. These examples set out details of compositions containing particular quantities of fipronil, organic solvents, crystallisation inhibitors, anti-oxidants and co-solvents which are said to have produced very effective products and which, upon application to flea infested dogs, resulted in a reduction of the flea population of over 95% as compared with a control group.
43. The independent claims as proposed to be amended and the various subsidiary claims for which independent validity was claimed were set out by the judge at [67]-[70]. I need only refer to the broadest of them, claim 1 as amended, which reads, in simplified form:

"Composition useful for treating and protecting pets infested or likely to be infested with parasites, characterised in that it comprises, in the form of a ready-to-use solution:

a) [fipronil];

b) a crystallisation inhibitor which meets [a glass plate test];

c) an organic solvent with a dielectric constant between 10 and 35, preferably between 20 and 30;

d) an organic cosolvent with a boiling point below 100°C, preferably below 80°C, and a dielectric constant between 10 and 40, preferably between 20 and 30;

wherein [fipronil] is present at the rate of 1 to 20% weight:volume in the composition."

The 564 patent

44. The 564 patent is entitled "Insecticidal combination effective against fleas on mammals, particularly cats and dogs".
45. The specification explains that the invention relates to a composition for controlling mammal fleas, for example fleas on cats and dogs, based on a synergistic combination of parasiticides which were already known.
46. It continues that fipronil and IGRs were both known but that the invention combines them together so as to provide parasiticides with both immediate and long lasting efficacy. More specifically, the invention is said to be a composition which includes fipronil and an IGR in a fluid vehicle suitable for local application as a spot-on and in which fipronil is present in a proportion of from 5-15% weight per unit volume and the IGR is present in a proportion of from 1-20% weight per unit volume.
47. On page 6 of the specification there appears a list of IGRs which includes all the IGRs which were known at the priority date and some which were not. This is followed on page 7 with a preferred list which includes a number of commercialised IGRs and some non-commercialised IGRs too. No explanation is given as to why they are said to be preferred.
48. On pages 9-10 the specification says that compositions for spot-on application can advantageously comprise a crystallisation inhibitor which is present in particular proportions and which satisfies the glass plate test; an organic solvent having particular characteristics; and an organic co-solvent, again with particular characteristics.
49. The specification describes the application of the combined product in the following terms on page 10:

"The compositions according to the invention intended for pets, preferably cats and dogs, are generally applied by being deposited onto the skin ("spot-on" application); this is generally a localized application over a surface area of less than 10 cm², especially of between 5 and 10 cm², preferably at two points and preferably localized between the animal's shoulders. Once deposited, the composition diffuses, preferably over the animal's entire body, and then dries without crystallizing or modifying the appearance (preferably absence of any whitish deposit or dusty appearance) or the feel of the fur."

50. There then appears a broad description of suitable organic solvents, crystallisation inhibitors, film-forming agents, surfactants and co-solvents in similar terms to that appearing in the 881 patent. But there is no further teaching as to how they are to be selected or combined save that, on page 13, the specification says:

"The compositions for spot-on application according to the invention are usually prepared by simple mixing of the constituents as defined earlier; advantageously, to begin with, the active material is mixed in the main solvent and the other ingredients or adjuvants are then added."

51. At page 14 the following tentative theory is advanced:

"The discovery that the compound (A), such as fipronil, dissolves in the sebum so as to cover the entire animal and becomes concentrated in the sebaceous glands, from which it is gradually released over a very long period, is a plausible explanation of this long-lasting efficacy for these compositions, and could perhaps also explain the long-lasting action of the associated compound (B) [the IGR]."

52. The specification concludes with some examples but, as the judge noted, and in marked contrast to the examples of the 881 patent, they simply specify different concentrations of active ingredients. They do not contain any formulation details save that there should be present a crystallisation inhibitor, an organic solvent and an organic co-solvent.
53. Claims 1 and 48 of the patent (as amended) and the various subsidiary claims for which independent validity was claimed were set out by the judge from [78] – [80]. Claim 1 is in the following form, again simplified as for the 881 patent:

A composition which provides small mammals with long-lasting protection against fleas, which includes, on the one hand, [fipronil]

and, on the other hand, at least one ovicidal compound (B), of insect growth regulator (IGR) type,

in a fluid vehicle which is acceptable to the mammal and suitable for local application on the skin;

wherein the fluid vehicle and the concentration of the [fipronil] and (B) are adapted for point application to the skin by deposition of the "spot-on" type;

wherein in the composition:

[fipronil] is present in a proportion of from 5 to 15 % (percentage as a weight per unit volume W/V); and

(B) is present in a proportion of from 1 to 20 % (percentage as a weight per unit volume W/V).

54. Omnipharm contend, and I accept, the other claims add nothing of substance. They do not include any claim requiring a co-solvent; they do not limit the crystallisation inhibitor to any particular compound or group of compounds; and in so far as they refer to solvents, they simply list standard solvent groups.

The judgment

55. This is a case in which the allegations of obviousness and insufficiency were closely related. Omnipharm argued that if the patents were not obvious then it followed they were insufficient. I must therefore outline the findings of the judge in relation to both issues.
56. The judge began by considering the obviousness of the 881 patent and identified the inventive concept of the broadest claim as being a spot-on formulation of fipronil containing certain excipients which is suitable for the treatment of pets infested or likely to be infested with parasites.
57. After dealing with various aspects of secondary factual evidence relied upon by both sides, the judge turned to consider the primary case, that of obviousness in the light of the Frontline spray. This was, of course, formulated in a manner suitable for application by spraying it all over the animal's body. The judge identified the differences between the Frontline spray and the formulation of the claim as being that the spray contains less than the claimed range of fipronil (0.25% rather than 1-20%) and that the spray does not disclose a crystallisation inhibitor or a co-solvent.
58. Omnipharm's case that it was obvious to take these steps and produce a spot-on formulation using the Frontline spray as a starting point ran as follows:

i) it was obvious to decide to try to develop a fipronil spot-on formulation given the success and efficacy of the Frontline spray and that spot-on formulations have an advantage in terms of ease of application;

ii) given that the Frontline spray functioned non-systemically, it was obvious to try, at least in the first instance, formulating a non-systemic spot-on; and

iii) the task of formulating a non-systemic spot-on formulation would be given to a formulator, who would be able to come up with a formulation within the claim without invention.

59. Merial challenged each aspect of this case. It argued that the available spot-on products, essentially permethrin and fenthion, did not have particularly good reputations. Further, if the skilled team contemplated a non-systemic spot-on formulation of fipronil they would think it would be worse than Frontline spray and they would have no scientific basis for supposing that it would distribute non-systemically from the point of application.
60. The judge rejected Merial's contention that the skilled team would not have an interest in making a spot-on formulation of fipronil and considered that the real question which would trouble the team would be whether such a formulation could be made to work. They would be faced with a choice as to whether to try to make a systemic or non-systemic spot-on formulation but would realise very quickly that there was a real difficulty in producing a systemic formulation in that fipronil has a log P of about 4.4 which makes it unsuitable for transdermal distribution. So they would have no alternative but to turn to non-systemic delivery. The key question then was whether it was obvious to try to make such a formulation with a fair expectation of success. The judge addressed this question at [124]:

"In my judgment, even taking account of the motivation to achieve a more convenient application method than a spray, the skilled team would not have a sufficient expectation of success to render the invention obvious. The team would have no common general knowledge basis on which to make such a prediction. The fact that other compounds had worked in spot-ons and pour-ons provides no technical basis for predicting that fipronil will do so, particularly when there is no scientific theory as to which properties of a compound, or characteristics of a formulation would make it do so. There is no evidence to suggest that the ability of a compound to distribute non-systemically is independent of its physico-chemical properties. The fact that fipronil is an efficacious compound in spray form does teach one that it works by contact with the insect's cuticle, and could in theory be made to work as a spot-on provided only that it will spread over the animal's body from the point of application. The skilled team would have no basis for predicting that that result could be achieved, or what formulation tools would be suitable for achieving that goal."

61. Importantly, the judge had regard to the lack of any general understanding as to the properties of a compound or characteristics of a formulation necessary to ensure its non-systemic distribution and concluded the team would have no basis for predicting which formulation tools would be suitable for achieving that end.
62. The judge then addressed the evidence of Dr Walters relied upon by Omnipharm in support of its obviousness case. This, the judge explained, was based upon Dr Walters' theory of dermal distribution. In his first report he said that the skilled team would be looking for solvents that would assist in spreading the formulation on the skin to give a large surface area from which its active ingredients could penetrate into the sebum and stratum corneum and then spread over the animal's skin surface. This, he said, would require a formulation with a low surface tension. He then continued at 5.72:

"As well as spreading, the formulator would consider techniques for ensuring that the active ingredient can penetrate into the sebum and surface layers of the stratum corneum at a rate which will encourage lateral diffusion. The active ingredient must, therefore, be kept in solution at a relatively high concentration but one that is not too close to saturation so as to avoid precipitation (including over the shelf-life of the typically two years). If precipitation is likely to occur, it will be necessary to add an anti-nucleant .."

63. The judge did not find this evidence persuasive, largely because Dr Walters was proposing a formulation which would cause the active ingredient to penetrate into the sebum and stratum corneum and then diffuse laterally. This would not, in his view, have been perceived as an easy or even a possible objective. Further, there was no general understanding as to how to use solvents for this purpose.
64. The judge also had in mind the evidence given by Dr Walters under cross-examination which revealed that he was contemplating a combination of a lower alcohol and transcutol, the alcohol to aid spreading and the transcutol to aid penetration. I think the passage in Dr Walters' cross-examination cited by the judge at [127] is very revealing:

"Q. And you say, "... as the alcohol evaporated, a concentration of the active ingredient in the main solvent would be formed that would be appropriate to facilitate penetration of the active into the sebum the outer layer ..." If you take fipronil as an example in a Transcutol ethanol system, which is what you are postulating, as I understand it -- yes?

A. Yes.

Q. -- as the ethanol evaporates, it will not leave fipronil at the high activity state in the residual Transcutol, will it?

A. No.

Q. Because it is less soluble than [in] ethanol?

A. Yes.

Q. How is it going to help?

A. You will notice I say "facilitate penetration", I do not say "enhance penetration".

Q. You say "facilitate" but not "enhance".

A. Yes.

Q. Could you please clarify for me.

A. I certainly will, although I thought I said this earlier in my testimony. What I am envisaging here is a formulation that will have the ethanol present to help the spread on the surface because it has a lower surface tension than the other materials and will facilitate the spreading of the liquid on the surface of the skin. As the ethanol evaporates, the fipronil will become less saturated in the residual solvent, which will be Transcutol or some other version of diethylene glycol monoethyl ether. In doing so the permeation from the residual solvent into the skin should be reduced. What I am trying to do is load the stratum corneum in such a way that I am not giving it that overload of the drug which will push it deeper into the stratum corneum. I am just letting the drug trickle into the stratum corneum which will give it a greater chance to diffuse laterally. Bear in mind that I have spread it over as wide as I can already with the ethanol and also with the help of the surfactant that I am putting in there."

65. The judge concluded that Dr Walters was putting forward an untested theory based upon a distribution mechanism which was not part of the common general knowledge. This, he held, was not a sound basis for an allegation of obviousness.
66. The allegation of obviousness against the 564 patent was necessarily weaker than that against the 881 patent because its claims are narrower, limited as they are to a combination of fipronil and IGR. The allegation having failed against the 881 patent, it followed that it must fail against the 564 patent too.
67. That brings me to the insufficiency attack. The judge addressed the allegation against the 881 patent at [134] – [139]. As I have said, it was advanced as a "squeeze", Omnipharm alleging that if the patent was not obvious then it must be insufficient because its disclosure is no more enabling than that of the common general knowledge and the prior art.
68. The judge thought this allegation had a short answer, namely that the specification contains many examples of formulations which were accepted to work as spot-on formulations for dogs and cats. The teaching of the specification therefore went further in a material respect than anything to be found in the Fipronil spray or the common general knowledge.
69. No doubt anticipating this response, Omnipharm developed its attack into an allegation that if it was not obvious how to make a spot-on formulation of fipronil from the prior art then the specification was not sufficient across its full width. It disclosed a series of examples which worked, without any guiding theory or principle to enable the skilled person to make anything but the examples actually disclosed.
70. In support of this allegation, Omnipharm relied upon this evidence given by Professor Hadgraft in cross-examination:

"Q. Well, presumably given that you have long lists of solvents, long lists of cosolvents, and long lists of surfactants, anti-nucleating agents, you would have to, as it were, consider, "Well, which ones from these lists am I going to use"?

A. You would have to make a judgment as to which ones you thought were most suitable.

Q. And how would you go about doing that?

A. I think that you go on and you find that there are specific examples given.

Q. So if you did not want to copy the examples, what would you do?

A. You would have to do some experiments to find out which ones were suitable and which ones were not."

71. It followed, submitted Omnipharm, Professor Hadgraft was unable to predict which formulations would work as a spot-on.
72. The judge was not impressed. He considered that the skilled team would have the teaching of the patent including, importantly, practical information about specific examples which worked. It had not been shown that this guidance, starting with the examples, failed to provide the skilled team with sufficient information to work the invention across the breadth of the claim. He put it this way at [139]:

"In my judgment the evidence does not establish that the claims of 881 are invalid for insufficiency. I must approach this issue on the basis of my findings that the skilled team would not have been able to make a fair prediction that they could make a spot-on formulation of fipronil starting only with the prior art and without the help of the patent. I must also approach the issue on the basis that the prior art provided no guiding principle for the selection of components of the formulation. On the other hand, the skilled team would have the patent in hand, together with practical information about specific examples which work. I do not regard the fact that the team would have to do some experiments based on the teaching of the specification and the examples to arrive at other spot-on formulations within the claim as adequately probative of insufficiency. It is true that the specification does not teach that the dermal theory of distribution is correct: but it does not have to. What is important is that it provides sufficient practical guidance to enable the skilled person to work the invention across its full breadth. It was not established that the guidance in the specification, starting with the examples, failed to do this. The allegation of insufficiency against 881 therefore fails."

73. Turning to the case of insufficiency against the 564 patent, this was rather stronger for two reasons. First, the skilled team now had to formulate a combination of active ingredients, not just fipronil. Second, the 564 patent contained no formulation examples at all. These differences were explored with Professor Hadgraft in his cross-examination in a passage cited by the judge at [150]:

"Q. Do you consider, assuming that you do not wish to copy the examples, that putting these two patents into practice would be a major research project with uncertain and unpredictable results?

A. I think because the patents give you guidance and they show that with examples that you can actually produce a product that works that it gives you confidence in them.

Q. So it is the examples that give you confidence?

A. Pardon?

Q. It is the examples.

A. If you have examples that you know work, then it proves that it is a possibility. It is possible to produce something that is commercially viable.

Q. In the '564 patent the examples do not give any formulations

A. '564 is the second?

Q. This is the patent we are looking at the moment.

A. But it gives you guidance that you need a solvent, a cosolvent and anti-nucleant agents.

Q. And that is it.

A. But it also gives you a restricted list of solvents and cosolvents and a restricted list of the anti-crystallisation materials.

Q. From which you would have to choose?

A. You would choose, yes. I mean, you cannot give a definitive answer because some of those anti-nucleants will be soluble in some solvents and not soluble in others and so you would have to find out or know personally which ones they are.

Q. Now, you have mentioned how challenging it would be to produce a formulation which combined two active ingredients like, as it were, fipronil plus an IGR, yes?

A. It would be significantly more challenging than just having one active.

Q. Yes."

74. The lack of any proper exemplification of the formulation of a spot-on involving the two actives was, the judge thought, fatal to the sufficiency of the 564 patent. He expressed his conclusions in this way:

"151. I think that the absence of any proper exemplification of the formulation of a spot-on involving the two actives is fatal to the sufficiency of 564. It is true that the specification teaches the need for a solvent, co-solvent and anti-nucleating agent. But that teaching is inadequate, in my judgment, to guide the skilled person to success. It is also true that the specification propounds the theory of dermal distribution, but I have held that the skilled person's general knowledge does not assist him to formulate a fipronil composition based on this theory. The bald proposition that the combination is efficacious undoubtedly goes beyond the teaching of the prior art, but is of no real practical assistance.

152. The rule of law that you "must not give people … problems and call them specifications" is as good now as when it was first uttered by Jessell MR in 1876 in Plimpton v Malcolmson 3 Ch D 531 at 576. I do not think that the very broad indication of the components of the formulation, either in the main claims or the subsidiary ones, is a sufficient description to enable the skilled person to arrive at a formulation within the claims without undue effort."

The insufficiency appeal

75. Mr Waugh submitted that the judge made several errors of principle in his approach to the issue of the sufficiency of the 564 patent. In summary, he contended:

i) the judge failed to take into account the whole of the teaching of the 564 patent;

ii) the judge erred in failing to recognise that there is no requirement for a patent to include specific examples;

iii) Omnipharm failed to adduce any evidence that the 564 patent could not be put into effect in any particular way and the judge ought to have found that it had failed to establish the patent is insufficient; and

iv) the judge's conclusion was inconsistent with the whole of the evidence.

i) the teaching of the 564 patent

76. Mr Waugh submitted the only matter the judge took into account was that the 564 patent teaches a solvent, a co-solvent and anti-nucleating agent. This, Mr Waugh said, is far too succinct a summary.
77. First, Mr Waugh pointed to the description of the active ingredients, namely fipronil and the IGR; the preferred IGRs; the ranges of concentration of each active ingredient; and the specific combinations of active ingredients set out in the examples, including their concentrations.
78. Second, Mr Waugh focused upon the disclosure of the nature of the various classes of excipients, namely the crystallisation inhibitor which passes the glass plate test; the organic solvent with a particular dielectric constant which must be present in a particular amount; and the organic co-solvent which must have a boiling point below 100°C, preferably below 80°C, a particular dielectric constant and be present in a particular co-solvent:solvent ratio. In this regard he also pointed to the teaching about the specific examples of excipients which may be used and the preferred examples of each category of excipient. He focused too upon the teaching that the formulations of the invention should be applied in a localised surface area of less than 10cm², preferably between the animal's shoulders and that after application the formulation diffuses over the animal's entire body. Further, he pointed to the teaching that the compositions of the invention are said to be particularly advantageous due to their high efficacy, their speed of action and the pleasant appearance of the animal's fur after application and drying.
79. Third, Mr Waugh drew our attention to the teaching of the results of the successful testing of the formulations of the invention on cats and dogs. He highlighted the unchallenged evidence of Professor Dryden that these results showed long lasting protection, that is to say that the formulations of the invention are very efficacious.
80. Overall, Mr Waugh submitted that the teaching of the 564 patent, including the particular preferred choices of the various excipients, provides detailed guidance to the skilled team and there was no evidence that any of the formulations of the invention, let alone the preferred possible combinations from the various lists of excipients, would not be effective and provide the promised long-lasting protection. Mr Waugh concluded that the judge failed to take nearly all of this teaching into account and this failure constituted an error of principle.
81. Attractively though these submissions were presented I found myself unable to accept them. As for the active ingredients, the judge referred at [72] to the fact the invention relates to a combination of fipronil and at least one IGR, and to the preferred IGRs including methoprene and pyriproxyfen. Moreover, in reciting the claim at [78], the judge plainly had in mind the disclosure of the ranges of concentration of each active ingredient. All of this information is, however, of limited value for the reasons given by Mr Watson in unchallenged evidence in his report, namely the list of IGRs includes all the generally known ones and more; no explanation is given for the choice of preferred IGRs, some of which Mr Watson did not even recognise; and the concentration ranges are so broad as not to be very helpful.
82. Turning to the excipients, the judge referred to the crystallisation inhibitors which pass the glass plate test at [66] in considering the 881 patent. But as Dr Walters said, the lists given in the patents include all the four groups of surfactants that were generally known as well as a number of other compounds. Moreover, the glass plate test could not serve as more than an initial screening tool; nor could it be used to predict how a formulation would behave when applied to the skin.
83. The disclosure in relation to solvents and co-solvents is something the judge also had well in mind, as is clear from [66]-[67]. Here he referred to the dielectric constant ranges which the solvent and co-solvent must meet, and that the co-solvent must have a boiling point below 100°C. He also referred to the lists of suitable solvents and co-solvents. This information would not, however, be of any practical use to the formulator who, as Dr Walters explained, would have to fall back on his general knowledge of solvents and techniques for enhancing spreading and skin penetration in order to decide on the appropriate solvent system to use.
84. The judge also took into account the teaching that the formulation should be applied to a localised surface area of less than 10cm², preferably between the animal's shoulders and quoted the teaching to this effect at [65]. So too he mentioned at [76] the disclosure that the compositions of the invention were extremely effective for the long lasting treatment of fleas on animals, and in particular small mammals such as dogs and cats. However, none of this provided any assistance on the key question as to how to formulate a combination product comprising fipronil and an IGR as a spot-on.
85. Finally, I must refer to the test results. This was a matter to which the judge expressly referred at [77]:

"77. In contrast to the examples in 881, the examples of 564 simply specify different concentrations of the active ingredients. The examples do not contain any formulation details beyond saying that there should be present a crystallisation inhibitor, an organic solvent and an organic co-solvent."

86. This then was the critical difference between the disclosures of the two patents. Omnipharm failed to establish that the practical guidance given by the examples of the 881 patent was not sufficient to enable the skilled team to work across the breadth of the claims. But the 564 patent claimed a combination of actives and did so without any worked examples at all. It provided no real practical assistance over and above the common general knowledge.

ii) The absence of examples

87. Here Mr Waugh focused upon the finding of the judge at [151] that the absence of any proper exemplification of the formulation of a spot-on involving the two active ingredients, that is to say fipronil and the IGR, is fatal to the sufficiency of the 564 patent. Mr Waugh submitted that two issues arise: first, there is no requirement in the Patents Act 1977 (or the EPC) for the provision of specific examples. Indeed, he drew our attention to two decisions of the Boards of Appeal of the European Patent Office, namely T 407/87 and T 561/96, which make it clear that the failure of the description to include any detailed example does not necessarily render the disclosure of an invention insufficient.
88. Moreover, Mr Waugh submitted that while it is correct there is no specific formulation example in the 564 patent, the examples on page 16 describe five different pairs of active ingredients which are said to be formulated in a liquid medium comprising a crystallisation inhibitor, a solvent and co-solvent.
89. I reject these submissions. I think it is clear from [151] – [152] that the judge did not find that the absence of any detailed examples was, in itself, fatal to the sufficiency of the 564 patent. What rendered it insufficient, in his view, was the absence of proper exemplification of a formulation of the invention in the context of a specification which was generally inadequate to guide the skilled person to success and provided no real practical assistance beyond the teaching of the prior art and the common general knowledge. The specification contains no more than a very broad indication of the components of the formulation and, as the judge found, it is not a sufficient description to enable the skilled person to arrive at formulations across the breadth of the claims without undue effort.
90. As for Mr Waugh's second point, the teaching on page 16 of the 564 patent does not advance the matter. Although it gives various different concentrations of the two active ingredients, it simply teaches that a representative solvent, co-solvent and anti-nucleating agent should be chosen. As I have explained, these were matters which the judge had well in mind.

iii) No evidence any formulations or classes of formulations did not work or could not be made

91. Mr Waugh contended that there was simply no evidence upon which the judge could come to the conclusion that the 564 patent is insufficient. Moreover, Mr Waugh continued, the judge found the teaching is inadequate to guide the skilled team to success, despite there being no evidence that anything falling within the scope of the claims does not in fact work. Further, the judge assumed against Merial that there was such evidence and, in so doing, in effect, reversed the burden of proof. This, Mr Waugh submitted, was clearly wrong and amounts to an error of principle.
92. I of course accept that the burden of establishing that a patent is insufficient lies upon the person making that allegation. As I shall explain in addressing Mr Waugh's next criticism, I do not, however, accept that the judge had no evidence upon which to reach his conclusion that the 564 patent was insufficient. Furthermore, the judge's conclusion was based upon his findings of non-obviousness both of the 564 patent and of the 881 patent. As the judge himself said in dealing with costs, there was, in relation to the 564 patent, a genuine squeeze.

iv) The conclusion was contrary to all the evidence

93. Mr Waugh recognised that a finding of insufficiency involves a multi-factorial evaluation by the judge of all the evidence against a statutory test which is expressed in simple terms, and that in such a case the finding of the judge must be treated with appropriate respect by an appellate court. For the reasons expressed by Lord Hoffmann in Biogen v Medeva [1997] RPC 1, we should be very cautious in differing from a judge's evaluation unless it can be shown that the judge has erred in principle or reached a conclusion which is perverse.
94. Nevertheless, Mr Waugh submitted that we should reverse the finding of the judge in this case because his conclusion was inconsistent with the evidence of the formulation experts, Dr Walters and Professor Hadgraft, both of whom thought that implementing the invention would be routine. Indeed, as I have said, Mr Waugh maintained the judge had no evidence upon which he could base a finding that the 564 patent placed an undue burden on the skilled team.
95. In addressing this submission it is convenient to begin with Dr Walters. It will be recalled that he believed that the skilled team would know of the theory of lateral diffusion through the stratum corneum, and that using this knowledge the team could readily produce a spot-on formulation containing fipronil.
96. Thus Dr Walters explained in his first report that the skilled team would be looking for solvents that would assist in spreading the formulation on the skin to give a large surface area from which the active ingredients could penetrate into the sebum and stratum corneum and spread from there over the whole of the animal's skin surface. Further, the team would consider techniques for ensuring that the active ingredients could penetrate into the sebum and stratum corneum at a rate which would encourage lateral diffusion and avoid precipitation. But importantly, Dr Walters thought the patents provided little guidance about to how to produce such a formulation above and beyond the common general knowledge. Specifically he said in relation to the 564 patent:

"9.10 But, as with EP 881, the formulator is left to himself on how to achieve a satisfactory formulation whether a spot-on or not. As regards a spot-on, as I said in respect of EP 881, dielectric constants are an unreliable guide to the solvents likely to be suitable for dissolving a given compound and the crystallisation test is unreliable as a guide to suitable surfactants or film-forming agents as crystallisation inhibitors. As I also said in paragraph 5.85 formulating combination products can be somewhat more challenging compared to single active products yet GB 564 gives no additional guidance for the formulator such as the physical properties of any of the listed IGRs. For these reasons, the skilled reader would have to rely on his pre-existing experience of formulation techniques and components and the common general knowledge to achieve a satisfactory formulation for a fipronil/IGR combination product."

97. In cross examination Merial challenged Dr Walters upon his evidence that the skilled person could readily produce formulations within the claims of both patents and hence they were obvious, but I do not understand there to have been any effective challenge to his evidence that the skilled team would have had to rely upon information that already formed part of their common general knowledge. Moreover, as the passage of Dr Walters' cross examination set out at [64] above shows, his opinions were founded upon his belief that the skilled team would have known of his theory of dermal distribution.
98. Ultimately the judge did not accept Dr Walters' evidence about the theory of dermal distribution, concluding that the theory did not form part of the common general knowledge. The judge also found that the skilled team would have had no other basis for predicting that fipronil, applied as a spot-on, would spread over the animal's body, and that the skilled team would not have known which formulation tools would be suitable for achieving that goal. It followed that Dr Walter's evidence that the 564 patent taught the skilled team no more about how to formulate such a product than they already knew from their common general knowledge became powerful evidence of insufficiency.
99. I turn next to Professor Hadgraft. In his reports he maintained it would not have been obvious to the skilled team in 1995 how to produce an effective formulation of fipronil which could be applied as a spot-on. The team would have had no understanding that distribution could occur by lateral diffusion and would not have thought it possible to produce a formulation that could be absorbed systemically. Overall, there were so many potential issues to consider that they would have regarded the task as a major research project with uncertain and unpredictable prospects of success. The task would have been even harder for a combination product comprising fipronil and an IGR. He put it this way in his first report:

"71. … Further complications arise because the second active will have different physicochemical properties. For example, the log (octanol-water partition coefficient) for methoprene is 5.5, compared with the value of ~4 for fipronil. It would be necessary to find a single set of conditions that provide adequate solubility, penetration, redistribution, stability etc for both actives at the same time. For example, the penetration of fipronil might be achieved using a particular solvent, but that solvent may not be appropriate for methoprene. It may also be that fipronil interacts with the IGR in an unpredictable way."

100. Then, in addressing insufficiency of the 881 patent, Professor Hadgraft explained that the specification gives specific examples of spot-on formulations. He continued that the 564 patent does not give specific examples but does give general guidance.
101. In his second report, Professor Hadgraft also made clear he did not accept Dr Walters' theory of dermal distribution:

" 6. … Dr Walters' approach is based on the premise that it would be possible to target the stratum corneum with fipronil, such that lateral spreading could be achieved across the whole of the animal in the stratum corneum layer. … there was no general understanding of, nor accepted scientific basis for, this approach. It was not one that I was aware of, nor was it an approach that was taught or understood. I do not consider that the skilled person would have thought that fipronil (or any active) could be effectively administered non-systemically via lateral spreading in the stratum corneum."

102. Then Professor Hadgraft developed his position on insufficiency, emphasising in relation to the 881 patent the importance of the examples:

"65. … the '881 patent discloses spot-on formulations of fipronil, which allow fipronil to be distributed over the entire body subsequent to application. It therefore does provide instructions and guides the choice of solvents and the crystallisation inhibitor, especially when read in conjunction with the examples. With reference to what Dr Walters says … regarding the examples, I do not agree that they give little guidance to the skilled person, who wanted to develop a spot-on formulation which is not exactly the same as the formulations described in the examples. Solvents and crystallisation inhibitors, chemically related to those mentioned in the examples, could be selected and, on the basis of the examples, be regarded as likely to provide an effective product. …"

103. But when it came to the 564 patent Professor Hadgraft could only say:

"85. … The '564 patent gives guidance as to what spot-on applications can advantageously comprise. The skilled reader does not have to rely solely on his pre-existing knowledge and experience."

104. Professor Hadgraft's cross examination is also illuminating. First, he adopted a more extreme position in relation to the theory of dermal distribution, as the judge recorded at [58]:

"As it emerged in cross examination Professor Hadgraft's view was more extreme. It was that some pesticide must first enter the animal's blood stream from the site of application and subsequently become concentrated at the surface, from whence it could contact the insect. He did not cite any support for this theory. … I do not consider that Professor Hadgraft's theory would represent the thinking of the skilled person. "

105. Professor Hadgraft was also asked about the teaching of the patents and, in giving the evidence set out at [70] above, emphasised again the importance of the examples in the 881 patent. But of course there are no such examples in the 564 patent. Here, as emerged from the further cross examination set out at [73] above, Professor Hadgraft accepted the skilled team had nothing but a general teaching of the theory of dermal distribution and the need for a solvent, a co-solvent and an anti-nucleating agent, of which lists are given. It followed that the skilled team would be left with the empirical task of trying to find out which combination of solvents and excipients would produce a formulation which worked. He also agreed that it would be significantly more challenging to produce a formulation containing fipronil and an IGR than a formulation containing fipronil alone.
106. So, in summary, the 564 patent provided the skilled team with a theory of dermal distribution and lists of solvents and excipients with which to set about the task of formulating the claimed combination of actives. But they would have had no understanding of that theory or which solvents and excipients should be employed to put it into practice, and no examples to assist them. As Professor Hadgraft himself said, it would be necessary to find a single set of conditions that provide adequate solubility, penetration, redistribution and stability for both actives at the same time. On the evidence before the judge, this was not a routine task and the 564 patent provides little by way of practical assistance. I am therefore satisfied that the judge had a proper evidential basis upon which to find the patent insufficient.
107. It follows it is not strictly necessary to address the further point raised by Omnipharm by way of respondent's notice but since we heard argument upon it, I will deal with it, albeit shortly. Omnipharm contended the claims of the 564 patent are not limited to any form of distribution and so include systemically acting spot-ons. However, given the judge's findings, the production of a systemically acting spot-on formulation of fipronil would plainly involve more than routine trials.
108. In my judgment this further point has no substance. The claims of the 564 patent (and, for that matter, the 881 patent) do not require a particular mechanism of distribution. They cover formulations useful for treating and protecting pets which have the particular features detailed in the claims and it matters not whether they act systemically or non-systemically. If the 564 patent had enabled the skilled team to produce the new and inventive formulations claimed without undue effort then it would have been sufficient.

The costs appeal

109. At the hearing of the appeal Merial sought permission to appeal against that part of the judge's costs order which required Merial to pay 40% of Omnipharm's costs, leaving aside the costs of various interim applications. That order was made after a further hearing before the judge on 27 January 2012.
110. The principles to be applied in considering an appeal against a final costs order were explained by Sir Murray Stuart Smith in Adamson v Halifax [2003] 1 WLR 60 at [16]:

"Costs are in the discretion of the trial judge and this court will only interfere with the exercise of that discretion on well-defined principles. As I said in Roache v News Group Newspapers Ltd [1998] EMLR, 161, 172:

"Before the court can interfere it must be shown that the judge has either erred in principle in his approach, or has left out of account, or taken into account, some feature that he should, or should not, have considered, or that his discretion is wholly wrong because the court is forced to the conclusion that he has not balanced the various factors fairly in the scale."

That statement was approved in AEI Rediffusion Music Ltd v Phonographic Performance Ltd [1999] 1 WLR 1507, 1523, per Lord Woolf MR. Although that decision was before the CPR came into force, it is clear that the court applied the same principle in relation to interfering with the trial judge's discretion."

111. In Islam v Ali [2003] EWCA Civ 612, Auld LJ expressed agreement with this formulation and gave the following further guidance at [19]:

"It is, as both counsel have acknowledged, a wide discretion, and the Court of Appeal should only interfere with the judge's exercise of it if he has "exceeded the generous ambit within which reasonable disagreement is possible", a familiar passage now taken from the judgment of Brooke LJ in Tanfern v Cameron-MacDonald (Practice Note) [2000] 1 WLR 1311, at paragraph 32, citing Lord Fraser in GVG (Miners: Custody Appeal) [1985] 1 WLR 647, 652."

112. Mr Waugh therefore accepted Merial had a high hurdle to overcome if it did not succeed on the main appeal. Nevertheless he submitted that the outcome of the judge's approach to costs meant that he had come to a result that failed fairly to reflect who had won and lost on the issues raised in the action and therefore who had won and lost on the issues in relation to which costs were incurred.
113. The costs consequences of the action fell to be considered against the commercial objectives of Omnipharm in bringing the proceedings and Merial in defending them and the findings which the judge made in relation to the major issues which arose for determination.
114. Mr Carr submitted, and I accept, that Omnipharm's commercial objective in these proceedings was to clear a path so that it could bring proposed formulations of fipronil on to the UK market without the risk of being sued by Merial for infringement of the 881 and 564 patents. On 28 May 2010 it served upon Merial four proposed confidential formulations designated A-D, each in two versions, one comprising fipronil as the only active ingredient and the other comprising fipronil with an IGR, together with a request for an acknowledgement of non-infringement. No acknowledgements were forthcoming and so Omnipharm commenced these proceedings.
115. In its defence, served some eight months before the trial of the action, Merial acknowledged that three of the formulations, A, C and D, did not infringe the 881 patent. However, the 564 patent remained a significant obstacle to Omnipharm because its broadest claims do not specify any excipients and protect any combination of fipronil and an IGR provided they are present in particular quantities. It was therefore necessary for Omnipharm to revoke the 564 patent to bring any combination product to the market.
116. I think it is fair to say that in their evidence for the costs hearing both sides gave emphasis to the importance of combination products. Mr Marshall, a partner in Taylor Wessing, the firm of solicitors acting for Merial, said in a witness statement dated 23 January 2012 that the UK market for Merial's combination product was £24m each year. Mr Cox, a partner in Fasken Martineau, the firm of solicitors acting for Omnipharm, explained in a witness statement dated 20 January 2012 that all the testing conducted by Omnipharm up to that point in time had been on its preferred combination product. Moreover it was, he said, Omnipharm's view that combination products would form the more valuable segment of the UK market in the next few years, reflecting the position in other major European markets, and Merial's marketing strategy being to switch customers away from mono products, for which there was generic competition, to its combination products, for which there was no generic competition.
117. The result of the proceedings was that the attack on the 881 patent failed entirely but Merial acknowledged that formulations A, C and D did not infringe it. The obviousness attack against the 564 patent failed but the insufficiency attack succeeded. The judge therefore ordered that the 564 patent be revoked, and made declarations in favour of Omnipharm that formulations A, C and D do not infringe the 881 patent.
118. In these circumstances Omnipharm contended that it had achieved its key commercial objectives and had cleared the path to enable it to market multiple fipronil spot-on formulations. It was now able to proceed with its plan to market in the UK three of its four proposed formulations including, so it said, its preferred formulation, both as mono products and, more importantly from a commercial perspective, as combination products. It therefore contended that it had won, but recognised there should be a deduction from its costs because it had lost on formulation B and on the validity of the 881 patent.
119. Merial contended that it was wholly successful in relation to the 881 patent which gives significant protection in respect of fipronil on its own or in combination with an IGR and, importantly, covers its commercial Frontline formulations. It is, in Mr Waugh's words, the jewel in Merial's crown. Moreover, what was at stake was not simply Merial's ability to stop one of the formulations proposed by Omnipharm, but to stop other parties who wished to sell other products within the scope of that patent.
120. Against this background it was hardly surprising that both Omnipharm and Merial each claimed that they had won and were therefore entitled to their costs, subject to some deduction to reflect the issues upon which they had lost.
121. In his judgment, given at the hearing on 27 January 2012, the judge reminded himself that CPR 44.3 sets out the various matters which should guide the court in exercising its discretion as to costs. He then identified the modern approach in patent cases as being (at [5]):

"… The court generally approaches the matter by asking itself three questions: (1), who has won; (2) has the wining party lost on an issue which is suitably circumscribed to deprive that party of the costs of that issue; and, (3), are the circumstances suitably exceptional to justify making a costs order on that issue against the party who has won overall."

122. In considering which party might fairly be said to have won, the judge cited the guidance given by Lord Bingham MR in Roache v News Group Newspapers [1998] EMLR 161 at 168:

"The judge must look closely at the facts of the particular case before him and ask: who, as a matter of substance and reality, has won? Has the plaintiff won anything of value which he could not have won without fighting the action through to a finish? Has the defendant substantially denied the plaintiff the prize which the plaintiff fought the action to win?"

123. After reciting the parties' various submissions, the judge identified a number of considerations which he believed relevant. First, Merial's costs appeared to be greater than those of Omnipharm by a ratio of about 3:2. Second, by far the largest issue at trial was that of obviousness of the 881 and 564 patents. Third, the issue of obviousness, upon which Omnipharm lost in relation to both patents, was clearly related to the issue of insufficiency upon which, in relation to the 564 patent, Omnipharm won. Indeed the judge observed at [2] that his finding that the 564 patent was insufficient was largely built upon his findings of non-obviousness of both patents. Later, at [16]-[17], the judge said:

"16. Fourthly, Mr. Carr rightly points out that obviousness of '881 and '564 on which he lost and insufficiency of '564 on which he won, are related issues. There was, in relation to '564, a genuine squeeze. It would be wrong therefore to suggest that, merely looking at the parts of the judgment which dealt with insufficiency of '564, will properly reflect the size of that issue as a whole.

17. In order to arrive at the position at which they ultimately arrived, Omnipharm would have had to have established that the skilled person would not have been able to make a formulation within the claims, something which was a fundamental part of their opponent's case on obviousness."

124. Recognising the complexity of the situation, the judge observed at [19] that when so much could be said on both sides there was a danger in the court starting from the position that one or other party must be the winner and that there are some cases which fall between these extremes and where, in truth, there is no overall winner. He continued that the present case was one where giving too much weight to a decision about the overall winner might cause an unjust result:

"20. I think the present case may present a different example of that principle where giving too much weight to a decision about the overall winner might cause an unjust result. I have borne that very strongly in mind in the approach which I have decided to adopt. It is of course tempting in such a situation to make no order as to costs at all, but I do not think it proper or fair to do that in this case."

125. The judge then came to his overall conclusions, finding first, that Omnipharm was indeed the overall winner:

"21. Standing back, it seems to me to be fair to view the matter in this way. Omnipharm has achieved success in relation to three formulations, but not in relation to formulation B and the validity of '881, which had great importance to Merial going far beyond this case. Moreover, invalidity of '881 would have given Omnipharm a significant further degree of freedom than that which they actually achieved. Nevertheless, if I apply Lord Bingham's approach and ask myself, has Omnipharm won anything of value which it could not have won without fighting the action and has Merial substantially denied the plaintiff the prize which they fought to win, I think I have to come to the conclusion that Omnipharm were the overall winners. I bear in mind also that the extent to which they have won is a factor which I can bear in mind in arriving at an appropriate proportion."

126. The judge decided this was not, however, a case for the application of a strict issue based approach. Instead, he considered it was appropriate to weigh all the matters to which I have referred and this he did in the following way in deciding that the justice of the case demanded that Merial pay 40% of Omnipharm's costs:

"22. It seems to me that this is not a case for the application of a strict issue based approach. For example, it seems to me that the fact that the issues of obviousness and insufficiency are so plainly interrelated does not justify refusing Merial a discount to Omnipharm's costs based on their success on obviousness, or indeed reflecting that in an order for costs to some extent going the other way. Merely deducting 15% or 20% of Omnipharm's costs does not seem to me to reflect the justice of the situation, nor by reflecting an order for costs the other way would twice that amount, 30% or 40%, be adequate, particularly when one bears in mind the greater amount of Merial's costs."

127. Mr Waugh submitted the judge made two errors in arriving at these conclusions. First, he wrongly decided that Omnipharm was the overall winner. Given that both sides were partially successful and partially unsuccessful, he ought to have concluded that there was no overall winner and he ought then to have examined the issues and determined who was successful on each of them before arriving at the appropriate overall order for costs.
128. Second, the judge's reasoning was founded upon a false premise, namely that the allegations of obviousness and insufficiency against the 564 patent constituted a genuine squeeze and that the finding of insufficiency was largely built upon the finding of non-obviousness.
129. Mr Waugh developed his first criticism of the judge's reasoning in the following way. He submitted that the court is not compelled to find a winner in situations where in fact there is not one. Further, in the context of this case, both sides achieved a good deal of success. Omnipharm revoked the 564 patent and secured particular declarations of non-infringement. Merial, on the other hand, successfully defended the 881 patent and a monopoly over its own commercial spot-on formulations of fipronil.
130. As for the judge's alleged second error, Mr Waugh submitted that the issues of obviousness and insufficiency were different. Omnipharm's case in relation to obviousness had three steps: first, it was obvious to decide to try and develop a fipronil spot-on formulation; second, given that the Frontline spray functioned non-systemically, it was obvious to try formulating a non-systemic spot-on; and third, the task of formulating a non-systemic spot-on of fipronil would be given to a formulator who would be able to come up with a formulation within claim 1 without invention.
131. Mr Waugh continued that it was the second of these steps upon which most effort was expended at trial and which ultimately proved decisive. Omnipharm failed to establish the necessary fair expectation of success and so its case failed. Further, this was an issue that simply did not arise in relation to the sufficiency of either patent because each taught that a spot-on formulation did work and provided experimental evidence demonstrating such efficacy. Moreover, Mr Waugh argued, although both obviousness and insufficiency raised in general terms the question of whether a formulation falling within the claims of the patents could be made without invention, the issues were in fact substantially different. The allegation of obviousness was founded upon the common general knowledge and whether the skilled person had an expectation that a spot-on formulation of fipronil could be made to work. The allegation of insufficiency, on the other hand, required a consideration of whether or not the skilled person could produce a spot-on formulation of fipronil without undue effort with the benefit of the detailed teaching of the specifications of each of the patents.
132. After hearing Mr Waugh develop his submissions, we decided it was appropriate to give Merial permission to appeal on a limited basis. We did not accept that it was arguable that the judge felt constrained, as a matter of law or principle, to find a winner. To the contrary, it is clear from [19] of his judgment that the judge was fully conscious that there are some cases in which there is no overall winner. Moreover, at [20], the judge recognised the present case as being one where "giving too much weight to a decision about the overall winner might cause an unjust result." However, we did think it arguable that, on the facts, the judge went further than was reasonably open to him in concluding that Omnipharm was the winner and that his overall costs award exceeded the generous ambit within which reasonable disagreement was possible. We therefore invited Mr Carr to respond on those issues.
133. Having heard full argument, I am satisfied that the judge was entitled to come to the overall conclusion which he did and that he approached the matter entirely properly. I say that for the following reasons.
134. First, Omnipharm has succeeded in clearing the way in respect of three of its four proposed formulations. Its particularly preferred formulation, on which it had done all of its development work, was a combination. It could not market this product without invalidating the 564 patent. Further, having invalidated the 564 patent, it is free to market not just its preferred product but other combination products too. There is no other generic fipronil combination spot-on product on the market and there is clearly considerable commercial value in being the first. In all these circumstances I think the judge was entitled to come to the conclusion that Omnipharm was the overall winner. It has won a substantial part of the prize for which it fought the action.
135. The judge's reasoning did not, however, stop there. He was fully conscious that Omnipharm failed in its claim for revocation of the 881 patent and this has particular importance to Merial, protecting as it does its own commercial product. As the judge recognised at [21], that importance extends beyond this case. It was with all these matters in mind that the judge explained that giving too much weight to a decision about the overall winner might cause an unjust result. Thus, in reaching his overall conclusion at [22], the judge settled on a figure which he thought fairly reflected the degree of success achieved by Omnipharm while also recognising the importance of the degree of success achieved by Merial.
136. Second, I of course recognise that the allegations of obviousness and insufficiency have different starting points. Omnipharm alleged it was obvious to formulate fipronil as a spot-on in the light of the common general knowledge and the success of the Frontline spray. By contrast, the insufficiency allegation was founded upon the teaching of the specifications of the 564 and 881 patents which Omnipharm contended was not adequate to enable the skilled team to implement the inventions of those patents without undue effort. In the context of this particular case there was, however, a substantial overlap between them because the judge concluded that the skilled person's general knowledge did not assist him to formulate a fipronil composition for application as a spot-on. This formed a key element of the judge's reasoning that the very broad indications in the 564 patent, amounting, in substance, to no more than the common general knowledge, was an insufficient description to enable the skilled person to arrive at a formulation within the claims of that patent without undue effort. The judge was therefore entitled to say, as he did at [2], that the finding of insufficiency of the 564 patent was largely built upon his findings of non-obviousness. Further, I believe he was entirely justified in saying, as he did at [16]-[17], that there was, in relation to the 564 patent, a genuine squeeze. The judge was satisfied that the skilled person was not able to make a formulation within the claims of the 564 patent based on his general knowledge and the generalised teaching and this was a fundamental part of Merial's case on obviousness.
137. When, however, it came to arriving at his ultimate costs award, the judge did not think that deducting 15-20% of Omnipharm's costs, that being the percentage by which the costs were increased by the allegation of obviousness, would reflect the justice of the case. Nor would twice that amount, namely 30% or 40%, be adequate, particularly bearing in mind the greater amount of Merial's costs. The justice of the case would only be met by applying a discount of 60% with the result that Merial had to pay 40% of Omnipharm's overall costs. This, it seems to me, was a conclusion to which the judge was perfectly entitled to come.
138. I would therefore dismiss Merial's appeal against the judge's costs order.

The maintenance of security pending the appeal

139. This is a very short point. Merial seeks the costs of applications made to Floyd J and to this court to preserve the security for costs provided by Omnipharm until after the hearing of this appeal. These costs were reserved to the hearing of this appeal by order of this court of 11 June 2012.
140. The background may be summarised as follows. Prior to the trial of the action Omnipharm provided security for Merial's costs in the total sum of £660,000. As I have said, judgment in the action was handed down on 21 December 2011 and a further hearing before the judge took place on 27 January 2012 to deal with the issue of costs and other consequential matters. At that hearing the judge made an order that the security be paid out to Omnipharm. He also refused an application by Merial for permission to appeal.
141. After the hearing but before the order was perfected and entered, Merial's solicitors, Taylor Wessing, wrote to Omnipharm's solicitors, Fasken Martineau, by letter dated 8 February 2012 requesting that the order for payment out of the security be stayed pending an application to this court for permission to appeal and, if the application was successful, pending its determination. On 10 February 2012 Fasken Martineau replied declining to agree to this request. Merial therefore made an application to Floyd J dated 16 February 2012 supported by a further witness statement of Mr Marshall dated 24 February 2012, and also raised the issue of security in its application to this court for permission to appeal and in its supporting skeleton argument. The witness statement of Mr Marshall was answered by a further witness statement of Mr Cox dated 29 February 2012 in which Omnipharm maintained its position.
142. The application duly came on for hearing before Floyd J on 2 March 2012. At the hearing Omnipharm offered a compromise, namely Fasken Martineau would undertake that, in the event this court granted Merial permission to appeal and ordered that some or all of the sums be maintained as security, they would pay the sums so ordered to be maintained back into court. This was acceptable to Merial and an order was duly made in those terms. Permission to appeal was subsequently given by this court and, since it appeared the appeal would come on for hearing relatively quickly, Omnipharm then agreed to continue the undertaking pending its determination.
143. I have to say that it seems tolerably clear to me that in all these circumstances Merial is entitled to an order for the costs of the applications for the maintenance of security it has made. It had to make the applications because Omnipharm declined to agree to the request it had made by letter of 8 February 2012. It was only at the hearing before Floyd J on 2 March 2012 that Omnipharm offered a satisfactory alternative.
144. Nevertheless, Mr Carr submitted that we should make no order for costs for the following three reasons. First, he submitted Omnipharm behaved reasonably throughout and, in particular, by offering a compromise. I accept that the compromise was a reasonable one to offer but in my judgment it could and should have been offered sooner, so avoiding the need for the further applications to Floyd J and to this court.
145. Second, Mr Carr argued that the costs were incurred as a result of Merial's own error in failing to raise the matter at the original hearing on 27 January 2012. I agree that the matter could have been raised at that hearing. But, as Mr Marshall frankly accepted in his witness statement, it was only after the hearing that Merial appreciated the possible effect of the payment out provision pending an appeal and it then promptly raised the matter in its solicitors' letter of 8 February 2012. It is only the costs after that date that it seeks and those costs could have been avoided by a positive response.
146. Finally, Mr Carr says the costs are relatively small. That may be so but they are nevertheless costs to which I believe Merial is properly entitled, although they should, I think, be offset against any order we make for costs in favour of Omnipharm in relation to the substantive appeal.

Conclusion

147. For the reasons I have given I would dismiss Merial's appeal against the judge's finding and consequential declaration that the 564 patent is invalid for insufficiency; dismiss Merial's appeal against the order requiring it to pay 40% of Omnipharm's costs; but order Omnipharm to pay Merial's costs of the applications to Floyd J and this court for orders that the sums paid into court be maintained as security pending the determination of this appeal.

Mr Justice Peter Smith:

148. I agree.

Lord Justice Richards:

149. I also agree.