England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
Lewison J.
HC03CO4495

		Royal Courts of Justice Strand, London, WC2A 2LL
		10/09/2004

Lord Justice Peter Gibson:

1. This is an appeal by the First Defendant, SmithKline Beecham plc, from the order made by Lewison J. on 25 April 2004 dismissing its application brought under CPR Part 11 and r. 3.1 for a stay of English proceedings brought by the First Claimant, DSM Anti-Infectives BV, and the Second Claimant, DSM Anti-Infectives Sweden AB, against the First Defendant and the Second Defendant, SmithKline Beecham Corp. The appeal is brought with the permission of the judge.

The facts

2. The First Claimant is a Dutch company which sells ingredients for pharmaceutical products. The Second Claimant is a Swedish company which manufactures such products in Sweden. I shall refer to the Claimants and to other companies in the group of companies of which they are members as "DSM". The First Defendant is an English company which manufactures ingredients for pharmaceutical products. The Second Claimant is a Pennsylvanian company which manufactures and sells pharmaceutical products using ingredients including those supplied by the First Defendant. The Defendants are subsidiaries of GlaxoSmithKline plc, an English company. I shall refer to the Defendants and to other companies in the group of companies of which they are members as "GSK".
3. Both GSK and DSM have developed strains of the bacterium Streptomyces Clavuligerus ("S.Clav."). They have used the bacterium to produce clavulanic acid which in turn is used to produce the potassium salt, potassium clavulanate. That salt is combined with a penicillin, amoxicillin, to manufacture an antibiotic, co-amoxiclav, which is more effective than amoxicillin in dealing with certain infections. GSK sell co-amoxiclav under the mark AUGMENTIN. DSM sell potassium clavulanate as a bulk product to generic pharmaceutical manufacturers. Each of DSM and GSK say that they have spent much time and money developing, through successive selective mutations of the bacterium, particular strains of S. Clav. to be as productive as possible of clavulanic acid. GSK's strain development work was carried out in England and its strains are held in the United Kingdom. GSK refer to their strain, which they isolated in 1981, as "SC7". DSM carried out their strain development work in the Netherlands, Sweden and Taiwan, where they used the services of a research company called Panlabs. DSM's strain bank is maintained in the Netherlands and clavulanic acid is produced by DSM in its plant in Sweden.
4. In 1988 GSK discovered that a competitor might have acquired a sample of SC7. They took the view that the sample was removed without authorisation by one of their own employees and sold to a Spanish company. They sued that employee and that company, and as a result of those proceedings they believed that all samples of SC7 had been destroyed. Later they learnt that SC7 had come into the hands of another competitor, a Slovenian company, LEK. They commenced proceedings against LEK and another company, Norton, which had intended to import and market in the United Kingdom co-amoxiclav manufactured by LEK.
5. On 23 October 1996 GSK's solicitors wrote to DSM, informing them of the proceedings against LEK and Norton, and saying that GSK alleged both patent infringements and misuse of confidential information in that they claimed that LEK used a strain of S. Clav. unlawfully obtained from DSM. The solicitors notified DSM that GSK would take similar action against DSM if their activities infringed GSK's rights. They reiterated that part of the action against LEK and Norton was for breach of confidential information and identified Dr. Callewaert as their former employee who had removed the SC7 sample. They also referred to Dr. Rowlands as another former employee who had been a leading member of GSK's strain development team but had become a senior employee at Panlabs. The solicitors said that GSK understood that DSM had had contracts with Dr. Callewaert and Panlabs in connection with DSM's clavulanic acid manufacturing activities and asked 4 questions relating to Panlabs, Dr. Rowlands and Dr. Callewaert, answers to which they sought within 14 days, and they requested DSM to provide GSK with information about DSM's co-amoxiclav product and the process for manufacturing it.
6. The letter was answered on 7 November 1996 by DSM which indicated that they were prepared to discuss both the matter of patent infringement and the matter of misuse of confidential information. GSK on 8 November repeated the request for answers to the 4 questions and said that failure to respond to that request would be taken to suggest misuse of confidential information.
7. On 14 November DSM wrote to GSK confirming DSM's preparedness to discuss both the matter of patent infringement and the source of any strain to be used by DSM. Although a meeting was arranged for 28 November GSK were not prepared to delay litigation, which they then commenced in the Netherlands on 26 November, alleging infringement of their patents by DSM. They did not allege in those proceedings misuse of confidential information.
8. After a number of meetings the parties entered into a settlement agreement dated 24 April 1997 ("the Agreement"). In the course of the meetings DSM disclosed to GSK details of DSM's development of DSM's strain of S. Clav. Those details were contained in documents which became Annex VII to the Agreement. In these proceedings DSM's strain has been called "the Annex VII strain" and DSM assert that that strain is derived from a publicly available wild strain, ATCC 27064.
9. The Agreement recited the proceedings brought by GSK against DSM for patent infringement, and GSK's opinion that DSM had infringed GSK's patents and DSM's opinion that they had not. It further recited DSM's use of the Diamine Ether process for the manufacture of potassium clavulanate and DSM's wish to continue to use that process, and that the parties desired to settle the proceedings. Recital (E) stated that the parties did not intend the Agreement to affect any of GSK's rights concerning potassium clavulanate in the USA.
10. The main points of the Agreement were the following:

(a) GSK acknowledged that the process which DSM had developed for processing the various strains, viz. the Diamine Ether process, did not infringe GSK's rights (cl. 2(a)).

(b) DSM agreed not to manufacture any potassium clavulanate using any process other than the Diamine Ether process or a process which did not infringe GSK's patent rights and agreed to deliver up to GSK for destruction all stocks of potassium clavulanate manufactured by DSM using the Disputed Process (cl. 2 (b)).

(c) GSK agreed that it should "not object to any future use, anywhere in the world, by DSM of any information generated by DSM through use of the Disputed Process" prior to the date of the Agreement (cl. 2 (e)). "Disputed Process" was defined in cl. 1 to mean "the processes used by [DSM] to manufacture Potassium Clavulanate which processes fall within the scope of the patents owned by [GSK] which are listed in Annex II". "Potassium Clavulanate" was defined in cl. 1 to mean the potassium salt of clavulanic acid, in bulk form, either alone, in combination with a diluent …., or in combination with amoxicillin …."

(d) By cl. 4 DSM agreed to certain obligations including temporary restrictions on sales of potassium clavulanate. The restrictions included in cl. 4 (a)(iii) agreement not to sell or supply any potassium clavulanate in any country in Western Europe in which there is a "Western European Patent". That term was defined in cl. 1 by reference to a list in Annex V. DSM acknowledged that any such sale or supply would be an infringement of the Western European Patent.

(e) By cl. 5 GSK agreed to certain obligations including in cl. 5 (i):

"[GSK] will not take action in any Court or Patent Office against [DSM] or any third party with respect to the Diamine Ether Process as developed by or for [DSM], or against [DSM's] use of the Disputed Process prior to the date of signature of this Agreement or the use by [DSM] of the strain with the genealogy and development details shown in Annex VII".

(f) Cl. 14 provides:

"EXCLUSION OF USA

Except only as otherwise specifically provided in Clause 2 (e) nothing herein contained shall be construed as a grant of any rights to GB under SB's rights with respect to Potassium Clavulanate in the United States of America. Further, SB expressly reserves all its rights with respect to Potassium Clavulanate in the United States of America and in particular its rights to defend and enforce such rights in the United States of America against any infringement."

(g) Cl. 15 provides:

"GOVERNING LAW

This Agreement shall be governed in all respects by the laws of England and exclusive jurisdiction with respect to all disputes in connection with this Agreement shall be given to the English Courts."

11. In late 2001 GSK developed a new test for identifying any use of SC7 in co-amoxiclav tablets. In February 2002 GSK tested tablets obtained in Austria from a German group, Hexal. They say that they discovered that those tablets contained both clavulanic acid produced from SC7 and diamine ether. This led GSK to believe that the clavulanic acid in those tablets originated from DSM.
12. In early 2002 GSK became aware that two generic pharmaceutical companies, Teva Pharmaceuticals Inc. ("Teva") and Ranbaxy Pharmaceuticals Inc. ("Ranbaxy"), both Delaware companies, were preparing to sell co-amoxiclav in the U.S.A. In August 2002 GSK commenced proceedings in the Philadelphia Court of Common Pleas in Pennsylvania against Teva and Ranbaxy, alleging that those companies were intending to sell co-amoxiclav derived from SC7 stolen from GSK. DSM were not originally joined in those proceedings, nor were they even mentioned, although GSK believed DSM to have been the suppliers to Teva and Ranbaxy of the potassium clavulanate. Teva and Ranbaxy challenged the jurisdiction of the Pennsylvania court, but that challenge failed in February 2003.
13. On 10 March 2003 Teva and Ranbaxy in response to interrogatories confirmed that DSM were their suppliers. On 15 May 2003 GSK obtained leave from the Pennsylvanian court to join two DSM companies as additional Defendants. One was served on 6 June and the other on 15 July 2003, but the service was technically defective and it was not until September 2003 that proper service was effected.
14. In the Pennsylvania proceedings GSK pleaded that in the course of the negotiations leading to the Agreement DSM had represented that the Annex VII strain was independently derived from strain ATCC 27064 and that GSK, on the basis of those representations, had agreed as a part of the settlement of the patent dispute in the Netherlands proceedings not to take any action against the use by DSM of the Annex VII strain. They claim against DSM (1) an injunction and damages for the misappropriation of a trade secret, (2) an injunction and damages for knowingly obtaining an unfair competitive advantage over GSK, (3) an injunction and damages for conversion of SC7, and (4) restitution for unjust enrichment and the imposition of a constructive trust.
15. On 7 October 2003 DSM disputed the jurisdiction of the Philadelphia court, but that issue has not yet been decided. It has been suggested to us that it may be that that court, which was made aware of the judge's decision on the construction of the Agreement, governed as it is by English law, and of the appeal to this court, is awaiting the outcome of this appeal.
16. GSK have a patent, falling within the definition of Western European Patent in the Agreement, extant in Italy. In 2003 GSK granted to an Italian company, Instituto Biochimico Italiano Giovanni Lorenzini SpA ("IBI"), a licence to manufacture co-amoxiclav in Italy for export. IBI had to obtain supplies of clavulanic acid from a third party. IBI approached DSM and informed GSK of this. On 8 October 2003 DSM wrote to GSK for confirmation that such supply by DSM would not be treated as a breach of cl. 4 (a)(iii) of the Agreement. GSK refused to give that confirmation, explaining by letter dated 19 December 2003 that they were alleging that DSM were using a misappropriated production strain of S. Clav. to manufacture clavulanic acid.

The English proceedings

17. On 23 December 2003 DSM commenced the present proceedings. They pleaded the Agreement, GSK's allegation that DSM was using stolen SC7 and the Pennsylvania proceedings. Para. 4. 6 of the Amended Particulars of Claim states:

"In fact, contrary to GSK's allegation, DSM has not been using SC7 as alleged by GSK or at all, but has been using at all material times either the Annex VII strains, as described in the Settlement Agreement and/or descendants of that strain produced by random mutation and selection techniques. All the strains used by DSM are or are derived from the strains described in Annex VII to the Settlement Agreement (and thus ultimately derived from a wild strain of Streptomyces clavuligerus deposited in a public culture collection as ATCC 27064). The strains were developed and selected as those identified to be optimal for the specific fermentation conditions developed in and for DSM's manufacturing process."

18. DSM further pleaded in paras. 6 and 7 the circumstances in which they say that a dispute has arisen over a supply by DSM of clavulanic acid to IBI. They pleaded in para. 7:

"Further GSK has asserted that the sale of clavulanic acid to IBI in Italy is contrary to Clause 4 (a) (iii) of the …. Agreement. If this is the construction of the clause then this would have the effect of preventing the sale of clavulanic acid irrespective of whether the sale of clavulanic acid by DSM infringed any of GSK's rights under the SPCs [Supplementary Protection Certificates] or could be reasonably understood to do so. Under such an interpretation the clause would constitute an unjustified and unlawful restriction on trade within the Common Market contrary to Article 81 of the EC Treaty. The clause should be construed so as to prohibit sales into Italy to the extent and only to the extent that such sales are an infringement of one of the SPCs."

19. DSM sought 10 declarations, viz:

"1.1 that DSM's potassium clavulanate has at all material times been manufactured using clavulanic acid produced from a strain of Streptomyces clavuligerus derived from those described in Annex VII of the Settlement Agreement and not from SC7;

1.2 that the representation by DSM BV that DSM BV was using a strain with the genealogy and development details shown in Annex VII to the Settlement Agreement was not misleading and false;

1.3 that GSK's strain SC7 whether by itself or generic material derived therefrom has not at any material time been used in the development of strains of Streptomyces clavuligerus used by DSM in the production of clavulanic acid;

1.4 that DSM has not used any alleged trade secrets incorporated in SC7 in the production in Sweden of its clavulanic acid;

1.5 that DSM will not infringe any trade secret rights in SC7 if it supplies its clavulanic acid made using DSM's strain of Streptomyces clavuligerus to IBI or any other country in the world;

1.6 that GSK is estopped from making any claim or objection in relation to the continued use of the strains used by DSM prior to the date of the Settlement Agreement;

1.7 that GSK is in breach of clause 2(e) of the Settlement Agreement by commencing the Philadelphia proceedings;

1.8 that GSK is contractually barred from commencing proceedings whether in the Commonwealth of Pennsylvania or elsewhere constituting objections to the use of the strains derived from those in Annex VII; and

1.9 that the Philadelphia Proceedings have been brought in breach of clause 15 of the Settlement Agreement.

1.10 that the sale of clavulanic acid to IBI is not prohibited by clause 4 (a) (iii) of the Settlement Agreement."

20. DSM also sought an injunction restraining GSK from taking further steps in, continuing or further prosecuting, the Philadelphia proceedings. However, they did not seek and have not sought an interim anti-suit injunction.
21. The First Defendant (the Second Defendant was not served with the proceedings until 14 June 2004) applied on 30 January 2004 for a stay of the English proceedings under CPR Part 11 on forum non conveniens grounds, viz. that the factual issues in dispute were already the subject of proceedings in Philadelphia which was the more appropriate forum for trial of the action. That application was heard by the judge. Before him GSK sought a stay permanently, alternatively until the Philadelphia court has determined the issue of jurisdiction, or in the further alternative until the European Court of Justice has ruled on a reference made to it by this court in Owusu v Jackson [2003] PIQR 186. I will explain that further alternative a little later.

The judgment below

22. The judge in his judgment first considered the scope of cl. 15 of the Agreement. He concluded that there was a dispute between GSK and DSM over whether DSM were using the Annex VII strain, that that was a dispute in connection with the Agreement and consequently that cl. 15 applied. Second, the judge held that any wrongful act by DSM must have taken place in Sweden where the confidential information was misused and that it was a misuse of cl. 14 to use it as a springboard for worldwide relief. He therefore approached the case on the basis that the exclusive jurisdiction clause applied to the subject matter of the dispute in the Philadelphia proceedings so far as it affected DSM.
23. Third, the judge considered a point taken by DSM on Art. 2 of Council Regulation (EC) 44/2001 on Jurisdiction and Enforcement of Judgments in Civil and Commercial Matters ("the Regulation"). Art. 2 provides:

"Subject to the Regulation, persons domiciled in a Member State shall, whatever their nationality, be sued in the courts of that Member State."

24. The exceptions provided for in the Regulation included in Art. 33 a provision that an agreement for the courts of a Member State to have jurisdiction to settle disputes confers exclusive jurisdiction on those courts. DSM relied on Art. 2 for a submission that the judge had no jurisdiction to grant a stay on the ground of forum non conveniens of English proceedings in which the First Defendant is an English company. The judge followed the decision of His Honour Judge Bentley Q.C. sitting as a judge of the High Court in Owusu. That was not a case on Art. 2 of the Regulation but on the like provision in the Brussels Convention. The judge refused to follow the decision of this court in Re Harrods (Buenos Aires) Ltd. [1992] Ch 72, viz. that it was not inconsistent with the Convention to stay proceedings on the ground of forum non conveniens in a case involving a conflict of jurisdiction between the English court and the courts of a non-contracting state. Judge Bentley held that he was obliged to refuse a stay in accordance with the principles which had subsequently been laid down by the European Court of Justice in Universal General Insurance Company v. Group Josi Reinsurance Co. [2001] QB 68 which cast doubt on whether the court of a Member State retains a discretion in such a case brought against a person domiciled in that state. On the appeal from Judge Bentley, this court referred to the European Court of Justice the question whether it was inconsistent with the Convention for a court of a Member State to exercise a discretionary power to decline to hear proceedings brought against a person domiciled in that state in favour of the courts of a non-contracting state. The hearing of the reference has now taken place but the opinion of the Advocate General and the decision of the European Court of Justice are still awaited. Lewison J. held that he had no jurisdiction to order even a temporary stay pending the ruling by the European Court of Justice in Owusu.
25. Fourth, the judge held that when considering an application for a stay under the court's case management powers, where there is an application to stay proceedings brought in accordance with an exclusive jurisdiction clause, there must be strong reasons for a court to grant a stay since it deprives the claimant of his contractual bargain.
26. Fifth, applying the principles applicable to applications for a stay on grounds of forum non conveniens, in view of the challenge to the Philadelphia court's jurisdiction and the judge's own view about the scope of cl. 15, he was not satisfied that there was another available forum having competent jurisdiction over DSM. The judge held that the factors on which GSK relied did not outweigh the force of the exclusive jurisdiction clause. One of those factors was the delay by DSM in commencing the English proceedings. The judge said in para. 37 (3):

"There has been some delay on both sides, it is true, but delay by DSM has not, in my judgment, been culpable. They took the jurisdiction point in accordance with the Pennsylvanian Procedural Code and more or less simultaneously began these proceedings."

A second factor was the potential risk of inconsistent findings. The judge in para. 37 (5) acknowledged that to be GSK's best point, but said that it had to be evaluated against GSK's claim that their testing was such that it would produce scientific certainty.

The appeal

27. The judge in granting permission to appeal acknowledged that his approach to the exercise of his discretion was "predicated on [his] construction" of cl. 14 and cl. 15 of the Agreement, that others might take a different view of those clauses, and if wrong on his construction his discretion might well have been vitiated.
28. On this appeal we have had the benefit of lucid and vigorous arguments from Mr. Jarvis Q.C. for GSK and Mr. Thorley Q.C. for DSM.
29. Mr. Jarvis makes the following submissions:

(1) The judge erred in his construction of cl. 14 and cl. 15: cl. 15 does not apply to the subject matter of the Pennsylvania proceedings, and cl. 14 entitles GSK to sue DSM in the Pennsylvania proceedings for committing actionable wrongs by selling products made with stolen SC7 to customers in the U.S.A.

(2) The judge incorrectly concluded that DSM had not culpably delayed bringing the English proceedings and that those proceedings were brought simultaneously with DSM taking objection to the jurisdiction of the Philadelphia court.

(3) The judge gave insufficient weight to the risk of inconsistent findings.

(4) The judge wrongly held that he had no jurisdiction to stay the English proceedings because of the Regulation.

(5) The judge wrongly thought that Pennsylvania was not an available forum, when it was the only forum in which all the disputes could be determined and the other relevant factors to be taken into account made it the relevant forum.

(6) The judge misdirected himself as to the correct approach to exercising the court's case management powers.

30. Mr. Thorley submits:

(1) there was no material error of law or principle underlying the decision of the judge, and that in the absence of such error, the exercise of discretion by the judge should stand without further question;

(2) the judge rightly construed cl. 14 and cl. 15;

(3) even if the judge was wrong on cl. 14, that does not affect DSM's right to pursue proceedings here;

(4) the judge made no error in finding no culpable delay by DSM and in holding that the risk of inconsistent findings was not determinative;

(5) England is the most appropriate forum;

(6) the judge made no error in refusing to exercise his case management powers to order a stay;

(7) in any event, because of the Regulation the judge was bound to hold that he had no jurisdiction to order a stay.

31. I start with the question whether the English proceedings are to litigate disputes or a dispute "in connection with" the Agreement. If they are, then DSM are exercising their contractual right under cl. 15 in bringing proceedings in England.
32. The central issue in the proceedings is whether DSM, as they claim, have been using the Annex VII strain, to which under cl. 5 (i) of the Agreement GSK can take no objection, or whether they have been using SC7, as GSK claim. Mr. Jarvis strongly contends that that is not a dispute in connection with the Agreement, because GSK accept that, if DSM have only been using the Annex VII strain and not SC7, DSM are entitled to do so. Indeed he goes further and says that, provided DSM do not use SC7, they are acting properly and have no need to rely on the Agreement. He calls it a mischaracterisation of the issue to label it a dispute in connection with the Agreement, which, as he points out, does not even mention SC7. He criticises the judge for accepting DSM's submission that in its context the Agreement must have been intended to settle all the disputes then pending between the parties, both for the alleged patent infringement and for the possible misuse of confidential information. He draws attention to the express references in the recitals to the Agreement to the patent infringement dispute and the absence of any reference to any dispute over confidential information. Insofar as DSM rely on cl. 2 (e) in addition to cl. 5 (i) as supporting their argument that the dispute about what strain they have been using is in connection with the Agreement, Mr. Jarvis submits that such reliance is misconceived because the strain used by DSM is not itself "information generated by use of the Disputed Process". Cl. 2 (e), he says, relates solely to information generated by use of the processes of manufacture of potassium clavulanate which are applied to the strain.
33. The judge did not rely on cl 2 (e), in my view rightly for the reasons given by Mr. Jarvis. But I am not able to accept Mr. Jarvis's other arguments relating to cl. 15. The prepositional phrase "in connection with" is one of wide import. If authority were need for so obvious a proposition, it can be found in Donohue v Armco [2002] 1 Lloyd's Rep 425 at p. 430 para. 14 where Lord Bingham said of an exclusive jurisdiction clause relating to "any dispute which may arise out of or in connection with this Agreement" that it was in wide terms, and that the practice of the English courts was to give such clauses, as between the parties to them, a generous interpretation.
34. Whilst the primary focus in the Agreement is on settling the disputes relating to the patent infringement alleged by GSK, it is plain from para. 5 (i) that the Agreement went further. The correspondence leading up to the Agreement, which I have summarised in paras. 4 – 7 above, shows that GSK were concerned about the possibility of a misuse of confidential information to produce the strain DSM were using, having regard to GSK's belief that Dr. Callewaert had stolen a sample of SC7, that DSM had been in contact with him and that DSM had been using Panlabs which employed Dr. Rowlands. In response to GSK's request for information DSM had made disclosure to GSK of the genealogy and development details of the Annex VII strain. The bargain struck between GSK and DSM involved each side accepting obligations and amongst those accepted by GSK was an obligation not to take action against the use by DSM of the Annex VII strain. In the Pennsylvania proceedings GSK themselves so plead, as I have noted in para. 14 above.
35. The question then arises: where a dispute has arisen as to whether in fact DSM have been using the Annex VII strain, as they assert, or SC7, as GSK assert, in what forum did the parties agree that dispute should be resolved, or was it left at large? In my judgment, subject to the effect of cl. 14 to which I shall come, the obvious answer is that it was not left at large but that such dispute is connected with the Agreement and so cl. 15 applies. It is connected with the Agreement because cl. 5 (i) bars GSK from objecting to the use of the Annex VII strain. True it is that that is a factual dispute, but I can see nothing to prevent such a dispute from being one in connection with the Agreement. The possibility of such a dispute arising was plain from the parties' exchanges prior to the Agreement.
36. Mr. Jarvis then says that cl. 14 is a proviso to cl. 15 and that because the dispute about the strain used by DSM relates to GSK's rights in the U.S.A. the dispute is taken outside the scope of cl. 15. He says that the words of cl. 14 are plain in giving unrestricted freedom to GSK to enforce their rights in the U.S.A. He says that the judge was wrong to accept that any wrongful act by DSM in relation to the misuse of confidential information can only have taken place in Sweden. He further criticises the judge for suggesting that GSK were seeking to use cl. 14 as a springboard for worldwide relief. He points out that in the Philadelphia proceedings GSK have limited their claims against DSM to the effects of DSM's allegedly wrongful acts in the U.S.A. and that whilst the injunctions they seek include one to restrain DSM from manufacturing potassium clavulanate using SC7, it is limited to such manufacture for import into the U.S.A.
37. Mr. Jarvis is, of course, right to say that torts may be committed in more than one country and that actionable wrongs against GSK can be committed in the U.S.A. arising out of the alleged misuse of SC7 in Sweden where DSM have their plant. However, I do not accept that cl. 14 prevents DSM from exercising their right under cl. 15 to bring the English proceedings to resolve the dispute in connection with the Agreement which has arisen even if cl. 14 allows GSK to bring the Pennsylvania proceedings. It is apparent from the evidence that GSK believe, as a result of the testing in 2002 of the Hexal co-amoxiclav tablets, that all DSM's potassium clavulanate uses the SC7 strain and is not limited to the Annex VII strain. Teva's and Ranbaxy's co-amoxiclav has not been tested by GSK, because it is not yet on the market. The same point is taken by GSK in relation to any supply by DSM to IBI.
38. Further, if it were necessary, which I doubt, for the purposes of this appeal to decide the difficult question whether cl. 14 does permit GSK to bring the Pennsylvania proceedings to determine the issue between the parties as to the strain used to produce clavulanic acid in Sweden and hence potassium clavulanate, I would hold that it does not. The wording of cl. 14 is significant. It reserves GSK's rights "with respect to Potassium Clavulanate in the [U.S.A.] and in particular its rights to defend and enforce such rights in the [U.S.A.] against any infringement". Those rights with respect to potassium clavulanate, as defined, are localised to being in the U.S.A. and are rights which might be infringed. That language suggests that the parties were contemplating patent rights, which are territorial, rather than the ownership by GSK of a strain of S. Clav. allegedly used in Sweden. As Mr. Thorley submits, the foundation of any claim which GSK may have must be DSM's production in Sweden, and any consequences in the U.S.A. are only secondary to DSM's activities in Europe and in particular Sweden. If GSK were to obtain against DSM the relief which they seek from the Philadelphia court, in reality they would obtain worldwide relief against DSM as DSM would be estopped from raising against GSK an inconsistent argument elsewhere. In my judgment therefore cl. 14 does not entitle GSK to have the issue of whether DSM have been using the SC7 strain rather than the Annex VII strain (neither side has suggested that DSM are using an unconnected third strain) determined in the U.S.A. in derogation of cl. 15.
39. In view of that conclusion I need say little on the less significant dispute over supplies to IBI. Mr. Jarvis submits that the pleading as to IBI and the declarations relating to it claimed by DSM are contrived. He describes DSM's argument that the sale of clavulanic acid to IBI would not be prohibited by cl. 4 (a)(iii) of the Agreement as "hopeless". I see very considerable force in Mr. Jarvis's submission, but I am not persuaded that we should hold that that point too, as pleaded in para. 6 and 7 of DSM's pleadings, is not a dispute in connection with the Agreement.
40. In the light of conclusions reached thus far, this court will not interfere with the exercise of the judge's discretion to refuse to grant a stay unless the judge can be shown to have erred in law or principle in according paramountcy to the exclusive jurisdiction clause. Mr. Jarvis suggested two such errors.
41. The first is the judge's findings in para. 38 (3) on the absence of culpable delay by DSM and the simultaneous commencement of proceedings. Mr. Jarvis emphasised that from the commencement of proceedings by GSK against Teva and Ranbaxy in August 2002 DSM must have been made aware that GSK were alleging that stolen SC7 was being used to produce co-amoxiclav, that from 6 June 2003 DSM were aware that they were being joined in those proceedings and that although the original service was invalid on the most technical of objections (the failure by GSK to serve the Agreement which DSM already had), that was made good in September, yet DSM did not commence the English proceedings until December 2003, and that in the meantime GSK had incurred $1.5 million in costs of which $225,000 has been spent on genetic sequencing tests by GSK's American expert, Dr. Sherman.
42. I accept that another judge might have taken a less charitable view than the judge of DSM's delay, but I am not persuaded that he made an error of law or principle such as vitiated his exercise of discretion. The action with which the judge was saying that the English proceedings were commenced "more or less simultaneously" was DSM's taking of the jurisdiction point on 7 October 2003. The English proceedings were commenced on 23 December 2003, and whilst it is somewhat generous to DSM to describe that as more or less simultaneous, the difference in time is not so great as to make the judge's view an error. It is clear that GSK and DSM have chosen throughout to act to their own tactical advantage, as is their right. GSK, despite their belief formed from the testing of the Hexal tablets in February 2002 that DSM were using SC7, took no action against DSM until the Pennsylvania proceedings and then did not seek to join DSM until May 2003. Because of GSK's procedural error DSM were not effectively served until September 2003, and DSM then promptly presented their jurisdictional objection. GSK have not provided a breakdown of the costs which they say they have incurred so that it could be seen at what times those costs were incurred. The judge has not, in my judgment, been shown to have erred in law or principle in rejecting the alleged prejudice to GSK from the delay as a weighty factor in the exercise of his discretion.
43. The second error said by Mr. Jarvis to have been made by the judge is the concern about inconsistent findings. Mr. Jarvis placed considerable emphasis on the undesirability of such inconsistency and he said that the judge failed to give it the weight it deserved. He also submitted that the judge, in referring to GSK's claim of scientific certainty produced by their testing, overlooked the fact that DSM disputed the accuracy of the test.
44. I am not persuaded that the judge made any error of law or principle on this point either. It is plain from the judge's assessment of this as GSK's best point that he did recognise that it was an important point. However matters of weight are for the judgment of the judge exercising his discretion and this court will rarely interfere with that judgment. He was entitled to take note of GSK's claim of the certainty which their testing would produce even though DSM challenged that claim. The reality is that the possibility of both the Pennsylvania and the English proceedings going ahead and producing inconsistent results is remote. Whichever of GSK and DSM lose in the first case to be tried on the issue of the strain DSM has been using would be estopped from arguing the contrary in the other proceedings. The case could and should be brought on for trial here quickly before a specialist patent judge. Although there would be no estoppel binding Teva and Ranbaxy or binding GSK as against Teva and Ranbaxy, it would be very surprising if the Pennsylvania proceedings were to go ahead against them on a contention for a different result, given the fact that the essential dispute is between GSK and DSM, and that Teva and Ranbaxy have no knowledge of the strain used by DSM other than from their suppliers, DSM.
45. In my judgment, therefore, it has not been shown that the judge made either of the alleged errors of law or principle. Nor has it been shown that on any other ground the judge has erred in a way that would justify the interference by this court with his exercise of discretion.
46. In the circumstances it is not necessary to say anything on the point relating to the Regulation save to record that Mr. Jonathan Hirst Q.C., sitting as a deputy judge of the Commercial Court in Travelers Casualty and Surety Company of Europe Ltd. v Sun Life Assurance Company of Canada (U.K.) Ltd. [2004] EWHC 1704, expressly did not follow the judge's decision on this. Any resolution of the difference of opinion should await a case where the court would be minded to grant a stay on forum non conveniens grounds.
47. For these reasons, I would dismiss this appeal.

Lord Justice Tuckey:

48. I agree.

Lord Justice Longmore:

49. I also agree.