Unilin Beheer BV v Berry Floor NV & Ors

If a patentee utterly prevails on infringement and validity and is held entitled to financial compensation in the Courts of England and Wales right up to the point where no further appeal lies, can all that be set at nought and utterly unravelled if the patent is later held invalid in the European Patent Office (“EPO”)? That is the main question we have to decide.

The essential facts (I try to leave out detail that does not matter) are these:

Unilin’s Patent involved was granted on 26^th June 2002 by the EPO. It is a European Patent (UK) (“EP (UK)”) (see below as to the meaning of this).

Unilin sued Berry Floor, IMC and B&Q (the “Defendants”) in the Patents County Court as soon as it was granted. Two types of product known as “Berry” and “Snap-fit” were alleged to infringe.

The claim included a claim for financial relief for infringements committed between the date of publication of the patent application (2^nd August 2000) and its grant. Such relief is provided for by s.69 of the Patents Act 1977 (“the Act”), implementing Art. 67 of the European Patent Convention (“EPC”).

In addition to defending the infringement claim the Defendants counterclaimed for an order for revocation of the Patent.

In September 2003 HHJ Fysh gave judgment for Unilin, holding claim 20 of the Patent valid and infringed by both the Berry and Snap-fit products, [2004] FSR 14.

This Court dismissed the Defendants’ appeal by a judgment and order of 30^th July 2004, [2005] FSR 6. The Court differed somewhat in its construction of the claim from that of the Judge below, but that made no difference to the result.

The House of Lords dismissed the Defendants’ Petition for leave to appeal in February 2005.

That left Judge Fysh’s order, and the order of this Court, in place. Besides an injunction (which by then did not matter commercially because the Defendants had changed to a non-infringing product) the key features for present purposes of the order as varied and the order of this court were that:

The reason for that stay lay in s.63(2) and (3) of the Act. This provides in effect that where a patent has been found only partially valid, neither damages (oddly not an account) nor costs can be awarded unless the patentee proves that the specification was framed in good faith and with reasonable skill and knowledge. Unilin proved the Patent was so framed, both at first instance and on appeal to this Court, [2006] FSR 495. The Defendants were ordered to pay the costs of the dispute about this, both at first instance and on appeal.

So Unilin then had unconditional final orders in its favour:

entitling it to proceed with an inquiry or account in respect of the Defendants’ dealings in the Berry products.

for 80% of the costs of the original trial and all the costs of the appeal, and

all of the costs of the issue of good faith and reasonable skill and knowledge both at first instance and on appeal.

All of those orders were the result of fully fought contests. No further appeal lies from any of them.

In November 2005 Unilin applied to HHJ Fysh for an interim payment as to costs. By consent he made an order for a payment of £500,000 with liberty to the Defendants to apply for an order for repayment of all or part of that following the ultimate decision in the EPO.

In March 2006 Unilin applied for directions for the damages inquiry or account of profits. The Defendants applied for a stay of those proceedings and of the assessment of all the costs orders. (Actually they had also made an earlier such application, but nothing turns on that). The basis of the Defendants’ application was that the “opposition” (as revocation proceedings are called) to the Patent in the EPO was not yet over and it might be that ultimately the patent would be revoked or limited in such a way that the Berry products would not be covered by it.

The rival applications were heard and determined by HHJ Judge Fysh by a judgment of 18^th May 2006. The Judge first had to consider whether Unilin’s entitlement to financial relief and costs was res judicata: if it was, as was common ground, there was no point in a stay. He held that there was no res judicata. That took him to the second point, whether or not he should exercise his discretion to order a stay. He decided to refuse one.

Following preliminary disclosure of figures Unilin elected to take an account. We are told the proceedings are substantial. Unilin contend that “profits by reason of the infringements” include not only profits from sales of the patented goods but also those from convoyed goods (e.g. underfloors), and “bridgehead” profits (i.e. profits made from the non-infringing products substituted for Berry but whose sales can at least in part be attributed to the fact that the business had been built up by Berry and Snap-fit). This will involve possibly difficult questions of causation. There is also said to be a problem of allocation of profits between those made here and those made in Belgium where the infringing products were made (Unilin are also suing in Belgium on the parallel Belgian patent). The hearing of the account is set for September this year.

Unilin appeal the finding of no res judicata, the Defendants the refusal to stay the inquiry or account.

For the uninitiated reader I shall briefly describe the current European patent system. I do not provide detailed chapter and verse. It is unnecessary and would be burdensome to reader and writer alike.

Until 1977, when the EPO, set up pursuant to the EPC, opened its doors, patents were purely creatures of national law. Each nation state had its own patent laws, patent office and judicial arrangements for the litigation of patents. There was no harmonisation of any of these – what had been achieved at international level was only the system for according priority to a patent application made in one state if an earlier application had been made within a year in another (the Paris Convention of 1883 as amended).

The idea of a European Patent and a European Patent system came out of the Council of Europe in the early 1950s. By the early 1960s there was substantial agreement as to what the substantive law should be, the agreement being embodied in the Strasbourg Convention 1963. Also the idea of a common patent office for Europe had come into being. Work on these two matters continued into the 1970s. In 1972 a Diplomatic Conference was convened. It led to the EPC in 1973.

The EPC is an international treaty between sovereign states. It is not part of the legislation of the European Union. Indeed a number of parties to the EPC are not members of the EU, Switzerland for example. As an international treaty it required implementation by national legislation. So far as the UK is concerned, that was by the Patents Act 1977.

Broadly the EPC did two things. First it prescribed what Member States should have as their substantive patent law, essentially following the Strasbourg agreement. Second it set up the EPO. This is a common office to which a patent application may be made seeking patent protection in the states chosen (“designated”) by the applicant. The application is processed and examined there.

If it survives examination a “European Patent” is granted. The term is a misnomer. The legal effect of a grant by the EPO is a series of national parallel patents each treated in each national law of the “designated” states as if it had been granted by the patent office of the state concerned.

Thus the 1977 Act did two main things, first it made substantive patent law accord with that laid down by the EPC and second it deemed a patent granted by the EPO with a UK designation (called a “European Patent (UK)”) to have the same effect as if it had been granted by the UK Patent Office pursuant to an application made to it.

Up until grant, the processing of an application for a European patent is essentially a matter between the applicant and the EPO. If the examiner objects to the application or part of it, the applicant has a right of appeal to a Board of Appeal, the EPC having provided for such an appeal and set up a system of Boards of Appeal. They are staffed by independent individuals – whose status I once described as “judges in all but name,” Lenzing [1997] RPC 245 at p.277.

In addition to provision of the Boards of Appeal, the EPC created an “Enlarged Board of Appeal”. Normally a decision of a Board of Appeal is final, there being no third tier of appeal provided for decisions about patentability. If however a difficult question of law arises, or if different Boards of Appeal differ on a question of law (decisions of law of one Board are not binding on subsequent Boards) then the Enlarged Board can be called upon, by reference from a Board of Appeal or in some circumstances at the request of the President of the EPO, to decide the matter. The parties themselves cannot appeal to an Enlarged Board, save (by a 2000 amendment of the EPC) in respect of an allegation of fundamental procedural impropriety.

Third parties have no direct say in the examination procedure. There is no means by which a third party can formally oppose grant. Those who created the EPC thought, almost certainly rightly, that if true opposition (i.e. opposition to actual grant) proceedings were available, patentees would be likely to suffer too much delay before they could actually enforce their patents. That had indeed had been the case in this country, for instance, under the opposition procedure of the 1949 Patents Act. As an alternative it would have been possible to leave all post-grant revocation proceedings to national courts. This potentially would have required (at least in theory) proceedings attacking the patent in all the designated states to clear the way for the whole of Europe. That was felt to be wasteful. On the other hand the parties to the EPC were not prepared to leave all questions of validity to the EPO. So a compromise was reached.

This compromise is that within 9 months of grant, third parties can make a central attack on the patent (by now strictly a bundle of national patents) in the EPO. The proceedings are called an “opposition” although they are in reality proceedings for revocation. The use of the word “opposition” is an indication that in some respects the proceedings are regarded as a continuation of the examination process. Perhaps for that reason the original examiner normally sits as part of the opposition division panel – judging his or her prior decision to grant the patent. Proceedings at this level are called “administrative” rather than “judicial” though it is difficult to see why – after all the opposition division is deciding a dispute between rival parties. Appeal lies to a Board of Appeal whose members, as I have said, clearly operate in a judicial capacity.

Apart from procedure by way of opposition however, the EPC also allowed attacks on validity (only for the state concerned, of course) in the national courts. Such an attack can be made from the date of grant right down to expiry and even beyond – you are not required to wait until an EPO opposition is over before you can attack a patent in a national court. And likewise the patentee can sue from the moment of grant, as indeed happened in the present case.

At the time of making of the Treaty it was thought that opposition proceedings would be relatively quick. But that has not proved to be so. Partly this is due to the success (in terms of numbers) of the EPO. There may be other reasons – for instance I have long wondered whether if the 9-month period were abolished, people would not bother to oppose unless they had to whereas now they have to get their opposition in often before they know whether the patent matters commercially. In the UK when the time limit for attacking a patent in the patent office was abolished in 1977, the result, contrary to what was expected, was an almost complete collapse of proceedings for revocation in the patent office. It may well also be that matters could be improved by way of procedure, for instance by tightening up on adjournments and avoiding as far as possible remission of cases from a Board of Appeal back to the Opposition Division as frequently happens now. Whatever the reasons, it is currently the case that opposition proceedings can take many years.

No one pretends that the compromise is satisfactory – it was a fudge at the time and remains so. Unless and until sensible judicial arrangements are put in place, the litigation of European patents in various national courts and the EPO will remain a messy, expensive and prolix business. One would hope that the politicians would find a way to put various national interests on one side for the sake of European industry as a whole. But despite attempt after attempt that has not yet been possible.

The result of the EPC compromise is that a national court system can find a patent valid by a final and conclusive decision and yet later, in opposition proceedings it is determined that the patent is invalid or must be reduced in scope. That is what may happen here.

We had some debate about this. Mr Henry Carr QC for the Defendants submitted that the intention behind the EPC was to make the EPO the “top” tribunal for validity. He took us to a sentence of my judgment in Lenzing at p.263:

He also relied on what was said by Laddie J in Unilever v Frisa [2000] FSR 708 at p.713:

Mr Carr also pointed to the fact that Laddie J seemed to have assumed that all costs incurred in fighting the action here might be wasted if there was a later revocation in the EPO. So he did. But that is no basis for thinking that he was in any way considering the estoppel point now before us.

Lenzing was about a challenge in our courts to an order of revocation made by an EPO Board of Appeal. I held that no such challenge could lie, even in the case of impropriety. Such an order is “final” in the sense that the patent is revoked and cannot be restored. (I would incidentally add that I am glad to see that by the EPC 2000 amendment impropriety of a Board of Appeal decision had been made the one class of case which can be taken to an Enlarged Board at the instance of the parties.)

As regards Unilever what Laddie J said was in the context of an application to stay UK proceedings. It was part of his discussion of whether the better course was to stay or not. He was not attempting to set out whether the national court or the EPO was supreme in some general sense. He was of course right to say that a subsequent ruling of invalidity in an EPO will take effect to revoke the patent even if a national court has previously held it valid. And he was right to assume that any injunction would go with the revocation – anticipating the Coflexip case discussed below. I note that he said nothing about what is to happen to financial relief granted by a final decision of the national court.

The other case cited by Mr Carr on this “who is top” point was Kimberly-Clark v P&G [2000] FSR 235. The main point of this case was a long way from the present problem. It was whether the previous UK rules about permission to amend a patent being a discretionary matter continued to apply to a EP (UK). The point arose because the patentee was applying in English proceedings to amend the patent and was simultaneously applying to amend the Patent in an EPO opposition. In the opposition proceedings as operated by the EPO there would be no discretion – so the patentee said there was no point in the English rules continuing to apply. This Court held that they did. But it also considered the problem of parallel proceedings in our courts and the EPO more generally. Aldous LJ said:

This does not help here. Of course in principle the preferred option is to stay UK proceedings if there are corresponding EPO proceedings. And it may in some circumstances be the case that an interim injunction could serve to hold the fort whilst these proceed. But all must depend on the circumstances and particularly the timing. Normally, although a stay is in principle the preferred course, it would be wrong to prevent the patentee from enforcing his patent here if the EPO opposition will not be concluded reasonably soon – as all too often it sadly is not. Take this case: the action started here in May 2002 and was finally over by November 2005. The EPO proceedings are still running and could be still doing so at the end of next year. Business needs to know where it stands – and a patentee is entitled to enforce his patent without undergoing the risks inherent on the cross-undertaking in damages – especially if the period involved could involve years.

None of these cases in my judgment assists in the debate. Nor does it help to ask whether a national court or the EPO is “top”. It all depends on the circumstances, as the two following scenarios illustrate:

The patent is still under opposition when a national court holds it valid and the EPO then revokes. So the EPO is “top”;

The EPO holds the patent valid and a national court subsequently revokes it (there is no estoppel created by an EPO decision as to validity, see Buehler v Chronos [1998] RPC 703). So the national court is “top.”

I turn to the proceedings in the EPO. The essential facts here can be stated much more briefly. In March 2003 Berry Floor’s parent company (“Berry Finance”) started opposition proceedings against the Patent in the EPO. This was close to the end of the nine month period from grant. The opposition proceedings are still not over – and are unlikely to be so for a year or more. It is possible that they will ultimately result in the Patent being revoked or limited so as not to cover the Berry products.

That is all one needs to know for present purposes. But it is worth perhaps briefly describing the points in issue in the EPO and what is happening. There are two arguments about the validity of the Patent.

The first arises from the fact that it was granted on the basis of a “divisional application” made pursuant to Art. 76 of the EPC. Where a patent applicant has made an earlier application which is still pending he may make a further application for just some of the subject-matter of that earlier application. Such an application is called a “divisional application.” A divisional application can be made on the initiative of the applicant alone. Or he may find he needs to make one as a result of the Office saying that his earlier application has got too much in it (is for more than one invention). Where the latter happens the applicant will have to “divide” out the inventions or just drop some of the subject-matter originally disclosed in the earlier application.

A divisional application will be “deemed to have been filed on the date of filing of the earlier application and shall enjoy any right of priority” (see Art.76). It would obviously be outrageous if the patentee could get a divisional patent where he had put new information (subject-matter) for the first time in his divisional application. If he could, he would get a date earlier than his actual disclosure of that subject-matter.

So what should happen if a patentee makes a divisional application which the EPO considers contains additional subject-matter? Clearly the Office cannot and will not allow the application. But can the applicant amend the application so as to delete the offending added matter? In the past the Office, pragmatically and sensibly said “yes”. But that practice was called into question by a Technical Board of Appeal in a case called Astropower T 39/05. The Board pointed out that there are problems, particularly where there is a divisional itself divided out of a divisional. The Board decided to refer a number of questions about divisional applications to the Enlarged Board of Appeal. The case reference is G1/05. Since then two more references raising further questions about divisionals have been made in cases G1/06 and G3/06. G3/06 was made by the Technical Board of Appeal dealing with the opposition to Unilin’s Patent the EP (UK) part of which was found valid and infringed here.

It is said that the divisional application leading to the Patent contained additional subject matter, that the application cannot be amended to delete it (even though the matter is, according to Unilin an irrelevant stray sentence) and that is the end of the Patent. Whether there really is additional subject-matter or if so, it is relevant, I do not know.

I say this, however. It would be pity if the Enlarged Board is forced to conclude that the prior practice was wrong. Patent application procedure should not contain traps of this sort, traps which cannot cured by amendment and where the cure would harm no one. After all if the patentee is allowed to cut out the offending matter he will get claims properly entitled to the date of the earlier application, claims supported by the description.

In passing it is worthy of note that the same problem was identified in the UK early on in the life of the new law. In Hydroaccoustics [1981] FSR 538 the Patents Court held that a divisional application made in the UK Office pursuant to s.76 of the 1977 as it then stood could not be cured if it contained additional matter. That was thought to be so unfortunate that s.76 was amended by the Copyright Designs and Patents Act 1988 so as explicitly to allow amendment to cure the problem. Whether, if the Enlarged Board says the prior practice was wrong, the UK Act is non-Treaty compliant and, if so whether that matters, is of no concern in this case but would obviously matter more generally.

The second point touching validity relates to priority. It is said that claim 20 (now claim 1 by amendment) is not entitled to the date of the first priority document and that, if priority is lost, the claim lacks novelty by reason of a prior use. Any cure by amendment of the claim would exclude the Berry products held to infringe.

This Court, like HHJ Fysh, held that claim 20 was entitled to priority from the first priority document. It is suggested that the Technical Board of Appeal, if the patent survives the decision of the Enlarged Board and is remitted to the Technical Board, will hold otherwise. I think that would be unfortunate. Some might say I would say that! But the point goes deeper. Coherence of the European Patent system requires that as far as possible different courts (whether parallel national courts or the EPO and national courts) should try to follow each other, particularly where a matter has been seriously contested. They should only differ where convinced the earlier decision was erroneous. Where the point is one where reasonable minds can differ, considerable deference to the first decision should be given. Of course this is not so if the cases turn on different points (e.g. different prior art) or different evidence. Then different results may well be reached without any inconsistency between tribunals. I say no more.

First I observe that the divisional point in the EPO, could, in principle, have been taken in the English proceedings, though it was not. There is no question of the Defendants in the English proceedings being bound here by a point they could never have taken here.

Second it was agreed that it does not matter that none of the Defendants are actually opponents in the EPO. The parties’ respective arguments for and against the res judicata point do not turn on whether there is identity between the party who unsuccessfully sought revocation here and the opponent in the EPO.

There are two important decisions of this Court, Poulton v Adjustable Cover [1908] 2 Ch 430 and Coflexip v Stolt (No 2) [2004] FSR 34. In Poulton the patentees had successfully sued the defendants for patent infringement, the court holding the patent to be valid and infringed. An inquiry as to damages was pursued. Before that had been determined the defendant discovered a further prior use. In a different capacity (as was possible then) it successfully brought revocation proceedings. It then unsuccessfully sought to avoid the damages ordered to be assessed in the first action. Both Parker J (a distinguished intellectual property judge) and the Court of Appeal (including another distinguished intellectual property judge, Fletcher-Moulton LJ) held that for the purposes of the inquiry as to damages the defendants were estopped by the judgment in the first action from denying the validity of the patent.

Poulton was re-considered recently by this Court in Coflexip. The facts were very similar to those of Poulton. In a first action the patent had been held valid and infringed and an inquiry as to damages was proceeding. Then, at the instance of a third party, relying on different prior art, the patent was revoked. The unsuccessful defendants in the first action sought to have the inquiry aborted by reason of that revocation. By a majority of this court, it was held that the defendant was estopped from denying validity in the inquiry.

An argument that things were different under the new patents law brought in by the 1977 Act was rejected by all three members of the court. An argument that the decisions of the House of Lords in Arnold v National Westminster Bank [1991] 2 AC 93 and Johnson v Gore Wood [2002] 2 AC 1 had made it open for this court not to be bound by Poulton was rejected by the majority (Peter Gibson LJ and Sir Martin Nourse, though it was accepted by Neuberger LJ). The majority also agreed with my reasons (given at first instance) for holding that Poulton should be good law as a matter of policy.

It is perhaps worth articulating those policy reasons a little more. First and foremost, the estopped defendant has had a full and fair opportunity of attacking the validity of the patent in his own proceedings. Next there is a very very strong public interest in the finality of litigation. Finally a party who had lost would have a strong motive for finding further or better reasons for attacking a patent and getting some third party to do so, thereby undermining the first decision. It is much better that he knows that the first litigation about validity is the time and place for him to get his best case together – that he knows he will have no second chance.

It might be suggested that there should be a different policy depending on whether the compensation has been finally worked out and paid, or, as in this case is still in the course of being worked out and not yet paid. Neuberger LJ thought in his dissenting judgment in Coflexip that once damages had been paid “there would be very substantial, possibly insuperable, difficulties in the way of Stolt getting those damages back”. I am bound to say that I think it very difficult to think of a rational ground for differentiating between cases where the compensation has been worked out and paid, and those where it remains to be worked out. I say that for several reasons:

Which of the two situations exists depends solely on the happenstance of the timing of the subsequent revocation. That date is in no way dependent on the speed with which the compensation assessment is carried out.

It is only as a matter of procedural convenience that the compensation is worked out after a finding of infringement of a valid patent. It could be worked out as part of the main trial. That certainly happened in at least one case in my practice at the Bar. It was a case of a pirate who had no defence to infringement and validity and the extent of whose dealings was known by reason of Norwich Pharmacal disclosure from Customs and Excise. The statement of claim not only alleged infringement but also quantified the damage at the same time and summary judgment including a fixed sum was obtained accordingly.

In Poulton Fletcher Moulton LJ specifically dealt with the point, saying:

It is not commercially sensible to draw the distinction. Once a man has a vested right to compensation he has a real asset – one he can deal in or raise money on. True it is not as good as cash, but it is commercially very different from merely the value of a cause of action.

Now a purist may say: it is a nonsense, and moreover an unjust nonsense, for a man to have to pay for doing what, with hindsight, we know to have been lawful. The purist might, I suppose, also say that a licensee who has paid royalties under a patent subsequently revoked ex tunc should get his money back. He might even say that a man who lost profits by refraining from some commercial activity by reason of a fear, now known to be groundless, of infringing the patent should have some remedy.

But I think there are good and pragmatic reasons why the purist approach makes bad business sense. You cannot unravel everything without creating uncertainty. And where a final decision has been made on a fair contest between the parties, that should stand as the final answer between them.

In a sense a patent is always potentially at risk – someone may come up with a bang on but obscure piece of prior art (my favourite pretend example is an anticipation written in Sanskrit wrongly placed in the children’s section of Alice Springs public library), or simply with better evidence on known prior art. That is no reason for undoing what has been done or regarding a final decision as merely provisional. After a final decision businessmen should be able to get on with their businesses, knowing what the position is.

In Coflexip there was a debate about whether the decision on infringement was an issue estoppel or a cause of action estoppel. The specifically pleaded instance of infringement related to the performance of a particular contract in which the patented apparatus and process were used. There were 14 other contracts whose performance involved the use of the same apparatus and process. The majority of the Court of Appeal thought that there was cause of action estoppel about the specifically pleaded contract but only issue estoppel about infringement so far as the other contracts were concerned (see p.748). They thought that the performance of each contract gave rise to a separate cause of action. This opinion was strictly unnecessary for the decision – for whichever type of estoppel applied did not matter. Either way the defendants could not challenge validity in the inquiry which covered all 15 contracts.

Mr Richard Arnold QC, sitting as a Deputy High Court Judge, in Hormel Foods v Antilles Landscape [2005] EWHC 13 (Ch), [2005] RPC 657 said he was “puzzled” by that conclusion. So am I, and not only for the reasons identified by Mr Arnold. Under the CPR a patentee must “give at least one example of each type of infringement alleged” (para. 11.1(1)(b) of the Practice Direction to Part 63). At trial the court will rule on whether each type pleaded is or is not an infringement. It will also rule on validity if there is a challenge to that. Under normal principles both rulings are res judicata and cannot be revised once all possibility of appeal is exhausted. And both rulings are not just on a sub-issue but on the causes of action decided upon at trial, namely whether or not the pleaded types of infringement do in fact and law infringe the Patent and whether or not that Patent is valid.

Now it is true that in an inquiry as to damages or account of profits the patentee is allowed to claim relief for types of alleged infringement not ruled on by the trial court. This saves the formal issuance of fresh proceedings in respect of these and is permitted as a matter of convenience, see General Tire v Firestone [1975] RPC 203 at p.207). And of course if, in the inquiry or account, the patentee alleges a type of infringement not considered by the trial court, the court conducting the inquiry or account will have to rule on whether it falls within the scope of the patent. But so far as the originally pleaded and proved type of infringement is concerned the matter is res judicata not merely as to an issue but as a cause of action in respect of that type of infringement.

Applying that to this case, it has not only been decided that the patent is valid, but also that the two types of infringement pleaded, “Snap-Fit” and “Berry” infringe. So, as Arden LJ pointed out in the course of argument, the cause of action for infringement has now merged with the judgment. At least for the infringements prior to the judgment (and there are none thereafter), Unilin simply no longer has any cause of action for infringement. It has a judgment instead.

It is common ground that both before and after the 1977 Act, an order for revocation of patent has retrospective effect. Going into Latin it is treated as revoked ab initio or ex tunc. Fletcher-Moulton LJ thought otherwise in Poulton but that is accepted to be wrong, see Coflexip at [137]. There is no provision in the Act expressly so stating, but since a successful attack on validity must be on the grounds that the conditions for grant were not satisfied, it follows inexorably that there can have been no power to grant the patent and hence it is a nullity.

Mr Carr accepts, as he must, that Coflexip and Poulton are binding on this Court. So, if Unilin’s patent had been granted by the UK Patent Office he would have to accept that the possibility of a subsequent revocation (it would have to be on the application of a third party) would not prevent the decision being res judicata as between Unilin and the Defendants. But, says Mr Carr, things are different because here we are concerned with a European Patent (UK).

The starting point for his argument – which involves an exception to the general rules of res judicata – must be the approach set out by Lord Diplock in Thrasyvoulou v Secretary of State for the Environment [1990] 2 AC 273 at 289:

So what Mr Carr needs are provisions in the Act from which it “can properly be inferred” that the doctrine of res judicata is to be excluded if a European Patent is revoked in EPO proceedings after a EP (UK) is held valid in our courts. I therefore turn to consider the relevant provisions. Because of their prolixity I set them out in an Appendix, focussing only on the key wording here. It was common ground that we can work on the provisions of the Act as amended. This includes the amendments made to take into account the implementation of the amendments to the EPC made in 2000. This is because, if in the event, Unilin’s patent is revoked in the EPO, that will not happen until the amended provisions have come into effect.

Part I of the Act sets out the provisions of the “New Domestic Law”. It begins with the requirements as to patentability, the domestic procedures for obtaining a patent through the UK Patent Office and the rules as to infringement. Some of the provisions implement those called for by the EPC but the whole of Part 1 is drafted as though there is no such thing as the EPO and no such thing as a EP (UK). It is Part II of the Act (“Provisions about International Conventions) and specifically ss.77-83 (“European Patents and Applications”) which makes patents granted by the EPO have effect within the UK.

Referring to the specific provisions s.130(1) defines the EPC, “European Patent” and “European patent (UK)”. The key provision which makes a EP (UK) have effect as a patent is s.77(1). Such a patent is to be treated “as if it were a patent in the UK under this Act granted in pursuance of an application made under this Act etc.” This is an unhelpful starting point for the inference which Mr Carr seeks. Both UK and European patents are to be treated the same. It is far from self-evident that a rule of res judicata which applies to a patent obtained through one route will not also apply to one obtained through the other.

Mr Carr disputes that. He says that the Act allows two routes of attack on a EP (UK). One is via the courts (or Comptroller) here, the other via opposition in the EPO. Only when both are over can one say that the patent is valid.

He says that a patentee who applies for a patent via the EPO route necessarily accepts that he may have his patent for a particular designated state revoked by either one of two routes, the national court or the EPO. Having made that claim he must accept the consequences. He seeks to get all that from s.74A and Art. 68 of the EPC. But all s.74A says is that where a European patent (UK) is revoked in accordance with the EPC, the patent shall be treated as having been revoked under the Act. That does not suggest in any way that an estoppel preventing a party from relying on a subsequent revocation is in any way different for patents obtained via the EPO route as opposed to those obtained via the national route. On the contrary it indicates the effect is the same.

Art. 68 provides for retrospectivity of revocation or limitation of scope in opposition proceedings. What it says is that “the resulting patent shall be deemed not to have had, from the outset, the effects specified in Arts. 64 and 67.” It is only necessary for present purposes to consider Art. 64, Art. 67 adding nothing to the debate. Art. 64(1) says in effect that a European patent for a particular state shall confer the same rights as would be conferred by a national patent in that state. Art. 64(2) adds nothing relevant for present purposes. Importantly Art. 64(3) says “Any infringement of a European patent shall be dealt with by national law.”

It is to be observed that Art. 68 has no equivalent express provision in the 1977 Act. Why not? For the obvious reason that it was not necessary to provide in UK law that revocation operated ex tunc. As I have said that is and always has been UK law anyway. The position may well have been different in other countries. The purpose of Art. 68 is to prescribe that the new parallel but national patent laws (I use the plural deliberately) required by the EPC should be the same in this regard. All should provide that revocation operates ex tunc.

Mr Carr submits he can get more out of Art. 68 however. He says “the effect specified in Art. 64” includes Art. 64(3), the bit about “infringement being dealt with by national law.” So, he submits, any effect of the national law of infringement is deemed not to have had effect if there is a subsequent revocation in opposition proceedings. It follows that any decision of a national court about infringement must be undone, along with all its consequences. This includes orders for costs, damages and so on. All, he submitted, must be undone. He felt some difficulty about the case where damages have not only been assessed but paid, but so far as I can see the logic of his argument must even include that case.

I reject all these arguments for a series of reasons. First is simply that Art. 68 has no equivalent in the 1977 Act. It is just part of an international convention. No doubt our Act must be interpreted as far as possible to conform to that convention, but there is nothing in our Act which suggests that our rules as to estoppel (which form part of our procedural law) are displaced by a revocation in an opposition.

Second even if Art. 68 were explicitly made part of our law, it would not help Mr Carr. For Art. 64(3) is not an “effect” at all. There is nothing which can be made retrospective about a provision simply saying infringement shall be dealt with by national law. It is only Art. 64(1) which provides for “effects” and so it is only those effects which are to be made retrospective.

Thirdly I cannot see how one can spell out of the Act – “infer” to use Lord Diplock’s word - an intention that something which our law regards as of (using Lord Diplock’s words) “fundamental importance” shall be excluded. Something as important as that would require much more. Putting it another way it is impossible to find in the Act any intention that a final and binding decision of our courts is to be regarded as only provisional – which is what Mr Carr’s argument inescapably involves.

Fourthly I do not think the untidy compromise reached by the Convention is that any final decision of a national court rejecting an invalidity attack is to be regarded as merely provisional. On the contrary the compromise was to allow a patentee to sue for infringement and for third parties to attack a national patent in a national court from the moment of grant. If the contracting parties had wanted to say that the result, if favourable to the patentee, is merely provisional, it surely would have so said in express terms.

Moreover if the result goes the other way, so there is revocation by the national court, the result is certainly not provisional: any subsequent decision in an opposition upholding the patent would not reinstate the patent in the State where the patent has been revoked by the national court. So if a national court’s finding of invalidity is final in the state concerned, why not its finding (as between the parties) of validity?

Fifthly there is the fact that by the time the court is considering damages or taking an account following a final decision that a particular type of product infringes a valid patent, by English law one is no longer even considering a cause of action for infringement. The underlying patent right has gone, being replaced by the judgment. So even if Art. 68 somehow retrospectively removes the cause of action for infringement which led to the judgment, that does not matter, one is no longer concerned as to whether a cause of action for infringement exists.

Apply that here. The account ordered by the order of HHJ Fysh of 26^th September 2003 is of “the profits that the First to Third Defendants have made by reason of their infringements of the Patent.” “The infringements” is in context a shorthand form for saying “sales and other dealings” of the “Snap-fit” and “Berry” products. You do not go back to the law of infringement at the stage of the account.

Sixthly, I consider that the travaux préparatoires to the EPC give a firm indication that national procedural law (which clearly would include the law as estoppel per rem judicatem) is to apply to European Patents when litigated in a national court. At our instigation the parties researched the travaux. The Defendants started at a very early stage. In 1962 there was a draft Convention relating to a European Patent Law. That provided for something very different than that which ultimately emerged, a Europe-wide patent with a European patent court dealing solely with validity and national courts dealing with infringement. The proposal in that form was not adopted. There is nothing in it which assists either side.

A new draft was formulated in 1970, abandoning both the idea of a Europe-wide patent and the idea of a central court for validity. Instead the idea was basically that which came to be, a central granting authority, whose grant of a patent in practice meant a bundle of national patents. Questions of infringement (including attacks on validity) were to be decided by national courts, though there was to be a “belated opposition” procedure in the proposed EPO.

A further draft convention emerged in 1971. Arts 18 and 105a of this were clearly the forerunners of Arts 64 and 68 of the EPC. They read as follows:

There was a report on the Conference. Each delegation provided a report on a different section of the draft convention. It was the British delegation which reported on Art. 105a. The report says:

This is valuable and to the point. The whole intention was not to interfere with the civil procedures of contacting states. To undo what we regard as of “fundamental importance” in our civil procedure would be a major interference with it – just what was not intended by the framers of the Convention. I do not agree with Mr Carr’s submission that this passage “lacks clarity” in any relevant respect.

The EPC was amended in 2000, the amendments taking effect at the end of this year. Mr Carr suggested that there was a change of intention by reason of the amendments to Art 68. But the amendment – to change the words “revoked in opposition proceedings” to “revoked or limited in opposition, limitation or revocation proceedings” – does not support the suggestion. The amendment is merely consequential upon the introduction of central limitation procedure in the EPO and by way of emphasis that a revocation by a national court is to have retrospective effect.

Mr Carr referred us to some cases which he suggested supported his arguments. I do not think any of them do, but I should explain why.

I have already explained why I think the cases about whether there is a “top” tribunal are of no assistance.

Next Mr Carr referred us to the Scottish case of ITP v Coflexip 19^th November 2004. There had been a trial in which the Lord Ordinary had held the patent valid and infringed. Whilst an appeal was pending a Board of Appeal of the EPO revoked the patent. The defenders withdrew all their grounds of appeal and instead were permitted to lodge an additional ground of appeal, namely that the patent was now revoked and to be regarded as void ab initio. Not surprisingly the Inner House allowed the appeal.

Mr Carr submitted that by implication the Inner House were here denying an estoppel created by a decision of validity. His technical argument ran thus: once the original grounds of appeal were abandoned, the decision at first instance became final. Yet the Inner House acted on the EPO revocation and allowed the appeal. So there was no estoppel created by the first instance decision.

The Lord President, giving the Judgment of the Court said:

The flaw in the argument is that it is obviously wrong to regard the first instance decision as “final”. It was not. It was under appeal - the very negation of it being final. As Mummery LJ said in the course of argument: “an appeal is just a continuation of existing proceedings. Res judicata does not get a look in.” The fact that certain grounds originally relied upon were no longer relied upon is not enough to make the decision below final. I think the Inner House would be astonished to be told that its decision had a bearing on the point we have to decide.

Much the same goes for the decision of this court in Reed Executive v Reed Business Information [2004] EWCA Civ 159 [2004] RPC 767. Certain parts of a first instance judgment were not appealed, other parts were. This court held that the unappealed portions could not create an estoppel in respect of any point raised on appeal. I put it a bit colourfully: “the appeal trumps the estoppel.” That is far from saying that a final, unappealable, decision can be trumped.

For those reasons I disagree with HHJ Fysh about these cases. He thought they supported Mr Carr. I do not.

Mr Carr also relied on the recent decision of this court in Special Effects v L’Oreal [2007] EWCA Civ 1. Special Effects sued L’Oreal on a registered trade mark. L’Oreal sought to contend that the registration was invalid. They had previously similarly contended in opposition proceedings in the Trade Marks Registry. It was argued that they were bound by that decision and could not subsequently defend the infringement proceedings by a challenge to validity. This court held that there was no estoppel arising by virtue of the decision in the opposition proceedings. Lloyd LJ said:

Anyone familiar with trade mark opposition proceedings will welcome this decision – for it is common experience that they are not generally fought with the tenacity (and expense) required when the parties are really at war with each other. Often the proceedings are really being used by both sides merely as a way of testing the water or providing a structure against which a settlement can be reached.

What I do not get out of the decision is any wide principle, such as that in any kind of case if a party has two routes of attack on the existence of a right a final decision by one route is never to be regarded as final. All depends on the provisions of the legislation concerned. Unless one concludes that those preclude an estoppel, under normal English rules an estoppel there will be.

Mr Roberts, appearing for B&Q, besides aligning himself with Mr Carr, took an additional point on the agreed form of the order of this Court. Following judgment the parties agreed an order which included a provision the material parts of which read:

Mr Roberts suggested this wording somehow precluded an estoppel even if without it there would be none. I do not agree. All the words do is to ensure that it is open to the Defendants to argue the point. They now have done so and have failed.

So I think the Judge was wrong to hold that there was no estoppel. I am not sorry to reach that conclusion. It means that businessmen in this country know that they can use the rather speedy court system here to get a conclusion one way or the other. If the patent is revoked, the way is cleared; if it is upheld and held infringed then compensation will be payable for past acts. And an injunction will run unless there is a later revocation by the EPO. Subject to that last point, the effect of all this is that one does not have to wait to find out who has won until the slowest horse in the race gets there.

That makes it unnecessary to consider any further points in detail. I will deal with them briefly. Firstly costs. Even if I had held otherwise in relation to res judicata I think the Judge was wrong to stay the costs orders. The plain fact is that even if the Patent is ultimately held invalid, Berry fought points on which they lost here. They should pay in any event. I note that even Neuberger LJ in his dissent in Coflexip thought that past costs orders could not be unravelled (see [34]).

Whether or not there should be a stay of proceedings for an account was primarily a matter for the discretion of the judge. Only if he is wrong in principle (taking something irrelevant into consideration or failing to consider something relevant) will this Court interfere.

I can see no error of principle by the Judge. He was clearly aware that an account might cost a lot of time and money. The choice was between keeping Unilin out of its compensation if they won in the EPO and the costs which would be wasted if there is no estoppel and Unilin lose in the EPO. In the latter event Berry can be compensated if they are wasted. At present Unilin have a judgment in their favour. Unless there are good reasons otherwise, they should be allowed to get on with enforcing it.

I therefore conclude:

That the Defendants are estopped from challenging Unilin’s entitlement to an account of profits, whatever the ultimate result in the EPO;

Even if there were no estoppel, past orders as to costs could not be challenged and so the liberty to the Defendants to apply for an order for repayment of costs should be discharged;

The Defendants’ appeal for a stay of proceedings relating to the account should be dismissed.

I agree with the judgment of Jacob LJ and would add the following small observations.

The strongest phrase in the European Patent Convention (EPC) that supports the contention of the defendants that revocation of the patent in opposition proceedings in the European Patent Office would affect the judgment already given in this action is the expression “as from the outset” in article 68. At first sight, these words appear to envisage that, once the patent has been revoked in opposition proceedings, the patent will be treated as having never had any existence. On that basis, there would be no basis on which a judgment in infringement proceedings could be maintained and enforced. The fact that article 68 does not provide that the effects specified in article 64 and 67 should be disapplied only with effect from the date on which the patent is revoked provides some support for the defendants’ argument (contrast for example Mason v Harris [1927] AC 252, a case on the dissolution of companies under the Companies (Consolidation) Act 1908).

In my judgment, however, there are a number of answers to this point. First, in so far as the EPC enables national courts to entertain proceedings relating to European patents, the national courts are given autonomy as regards procedure (see article 64(3)). The effect of the prior judgment for infringement is thus a matter for national law. Secondly, as Jacob LJ has recorded, once the cause of action for infringement is given, it is as a matter of English law merged into the judgment and the owner of the patent becomes entitled to a new piece of property, namely the judgment. As I see it there is nothing in the EPC to suggest that revocation is to take away this property right.

Thirdly, it is a general principle of the rule of law that legislation should not have retrospective effect without good reason. There may be said to be a presumption against retrospectivity. It is difficult to see any reason for treating article 68 as retrospective in the factual situation in this case, and accordingly this court should not give it that effect.

For these reasons, and those given by Jacob LJ in his judgment which I have had the benefit of reading in draft, I would make the order proposed in paragraph 92 above.

I agree with both judgments. I should add that after they had been written the parties informed us that they had settled their dispute and were proposing to file a consent order providing for service by each side of notices of discontinuance of the claim and counterclaim respectively. At present we are still seized of the appeal. This judgment is being delivered with the consent of both sides, though the order proposed will not in fact be drawn up.