Validity and Infringement: a surrealist distinction for the Belgian patent courts


1.   Belgian patent law professes a system where validity and infringement of a patent should be dealt with simultaneously. Like the neighbouring countries of France, the United Kingdom, the Netherlands and Luxembourg, in Belgium there is no special court or patent office with exclusive jurisdiction to hear claims for patent revocation. When an infringement claim is brought before one of the five specially designated patent courts, the defendant may bring a counterclaim for invalidity and the court will assess the infringement of the patent simultaneously with the issue of validity. Since an invalid patent cannot be infringed, the court will start with a determination of the patent’s validity and will only afterwards deal with its infringement.

2.   There are several advantages to such a “combined” system. In the first place, as stated above, it allows a patent to be “tested” for validity before being asserted for infringement. In the second place, it prevents the patentee adopting different positions in relation to validity on the one hand and the infringement on the other. Such an approach has been likened to the behaviour of an Angora cat: “When validity is challenged, the patentee says his patent is very small: the cat with its fur smoothed down, cuddly and sleepy. But when the patentee goes on the attack, the fur bristles, the cat is twice the size with teeth bared and eyes ablaze” [1]. It is believed that in a “combined” system such tactics are excluded. As the English judges observe, “it is desirable to try infringement and validity issues together, where at all possible. If they are tried separately it is all too easy for the patentee to argue for a narrow interpretation of his claim when defending it but an expansive interpretation when asserting infringement.” [2]

A “combined” system therefore allows arguments for validity and infringement to be “squeezed” together, ensuring that they are mutually consistent. In a typical “squeeze” an alleged infringer will use the patentee’s arguments for overcoming the prior art for his own benefit: either to show inconsistency between them, or to show that the patent is not infringed for the same reasons as the patent was found to be novel over the prior art.[3]

3.   This “combined” system is however not as universal as one would think. Many European countries, most notably Germany but also Poland, Austria, Slovakia and the Czech Republic, have legal systems in which the validity of a patent is dealt with separately from the courts. Whereas the proposed Community Patent Regulation provides for courts that will, in principle, hear validity and infringement simultaneously, it also provides for the possibility of “bifurcating” these issues, in that the question of patent validity can be referred to the Central Division while the infringement question will remain with the originally seized infringement court.[4]

4.   The purpose of this paper is not to give an overview of all these systems, nor to take a position as to which system (“combined” or “dual”) ought to be given preference. This paper is intended to illustrate how in a country where patent litigation had been slow and inefficient for many years for reasons undoubtedly related to the “combined” system, the Belgian courts have substantially improved both the quantity and quality of their decisions and have shed their reputation for being one of the slowest courts on the continent by separating the question of validity and infringement in appropriate cases.

It will be shown below how Belgian patent litigation practice gradually evolved into a system where bifurcation is often deliberately chosen by the courts or the parties themselves to overcome the perceived deficiencies in the litigation system, or simply for legal and strategic reasons.

 A.  How the presumption of prima facie validity became the legal basis for granting preliminary injunctions

5.  The main reason why the Belgian courts were so notoriously slow in handling patent disputes, especially during the 1990’s, was that these disputes were systematically presented to them as so complex that a civil judge would be unable to solve all the questions on his own, but rather would require the assistance of specialized court experts. Bringing infringement and validity issues together was therefore not particularly helpful for speeding up or streamlining the procedure. The long delays caused dissatisfaction on the parts of all parties involved.

The first step: ex parte seizure actions

6.  One particular procedure brought about a change to this situation. The well known ex parte “saisie description” allowed a patentee to obtain and enforce a discovery measure from a Belgian court in a matter of days. This procedure was designed in the first place to obtain, via a court-appointed expert, irrefutable evidence of infringement so that the patentee could decide whether or not (and where) to launch infringement proceedings. This procedure evolved during the nineties when the demand for fast and efficient patent enforcement measures was increasing, and was the only protection for patentees who could not afford to wait for the outcome of the main proceedings. In addition to seeking evidence of infringement, the procedure also allowed the patentee, by way of a preliminary measure, to put seals on allegedly infringing products or to request an injunction to prevent their movement or disposal. Whereas this “injunctive” measure was originally used for making sure that the evidence was preserved during the discovery period, the Belgian courts gradually became more and more willing to expand the reach of this measure beyond the time needed by the expert to collect evidence and, increasingly, until a final decision had been reached in the main proceedings, thus effectively granting a true “ex parte injunction”. In practice, this meant that the preliminary measure remained in force for a few years, so that the focus and efforts in a patent dispute shifted from the main proceedings towards the interim measures that could be obtained.  

 7.   How could these seizure judges grant preliminary injunctions that lasted for a few years, when the regular courts were only prepared to grant permanent injunctions after a full (and long) examination of patent validity and infringement? Because the seizure judges took quite a liberal approach towards the issue of patent validity. Indeed, relying on their long standing practice in more “classical” seizure cases, they considered that a patent was an official title granted by a public authority, and as such ought not to be challenged for the purpose of their proceedings. This reasoning was inter alia supported by the fact that around the same time the examination practice of the European Patent Office (EPO) had gained sufficient authority to convince the Belgian judges that if the EPO had granted a patent after a substantive examination in relation to novelty and inventive step, the patent ought to be taken seriously and not easily revoked. This argument could not be made for patents granted by way of “rubber stamping” at the Belgian Patent Office, where no substantive examination was carried out. The combination of these factors explains why Belgian judges developed the legal presumption of “prima facie” validity of a European patent. This presumption allowed a patentee and the seizure judge to disregard all discussions about patent (in)validity and consider only issues of infringement. Hence the proceedings before these courts became “bifurcated” in that the patentee had to show only that he held a patent that was granted by the EPO and had been validated in Belgium. His burden of proof was then limited to showing a prima facie case on infringement and the reasonable nature of the requested measures. [5]

8.  This concept of prima facie validity, which was inspired by principles of Belgian administrative law, had another far reaching consequence. It allowed a seizure judge to accept that a European patent was prima facie valid even when it had been revoked by the (first instance) Opposition Division of the EPO if that decision was under appeal before the Technical Board of Appeal. This was because article 106.1° of the European Patent Convention provides that an appeal against such a decision has a “suspensive effect”. Hence, the Belgian seizure judges reasoned that the title as originally granted continues to exist and the patent should therefore remain enforceable for preliminary injunction purposes.

These principles were established by the Antwerp appeal court in its landmark “Epilady” decision of 12 June 1991 and have subsequently been confirmed by several decisions of the Belgian appeal courts.[6]

The second step: inter partes “kort geding” actions

9.   Another development arising at the dawn of the 21st century was that Belgian courts became increasingly prepared to entertain patent disputes via emergency proceedings (known as “kort geding”, as in Holland), where the principle of prima facie validity was likewise adopted. Since main proceedings generally took several years to complete and the outcome of EPO opposition proceedings was not considered relevant until a final decision had issued, it became easier for a patentee to demonstrate the urgency of his case, even if the “kort geding” action were to take four to six months to get to trial. As a result, more and more patent disputes were conducted via inter partes emergency proceedings, allowing patentees to obtain some relief within a reasonable timeframe. Instead of trying to anticipate what a court would do in the main proceedings (the applicable legal standard in the Netherlands and other jurisdictions) the Belgian “kort geding” judges gratefully adopted the principle of the prima facie validity of a European patent, as developed by their colleagues in the “seizure” courts. By applying this bifurcation, the courts conveniently avoided having to address the complex issues of (in)validity and were able to decide cases more quickly. This of course denied the alleged infringer the “defence” of invalidity and thus created a very pro-patentee climate. Many patentees chose to “try their luck” in Belgium because of the ease with which injunctions could be obtained and the fact that security was (almost) never requested. In relation to pharmaceutical litigation against generic drugs where only patent validity was in dispute, this development explains why many generic drugs were unable to find their way onto the Belgian market before the EPO opposition or the main Belgian proceedings had come to an end.   

The third step: cross border preliminary injunction actions

10.   Inspired by the Dutch cross border injunction practice, the presumption of prima facie validity of a European patent offered a basis to the Belgian courts for granting pan-European patent injunctions. The first one was granted as early as 1999, but remained unnoticed in the literature because both parties were Belgian and a purely “Belgian” injunction would have given the same effect.[7] But then, on 14 September 2001, the Brussels first instance court issued its first “genuine” cross border decision, when the Unilever group of companies, with seats throughout Europe, was subjected to a preliminary injunction.[8] In 2005, that same Brussels court of first instance issued a cross border injunction against a series of generic drugs manufacturers from India and Pakistan, who had presented their products at a Brussels trade fair.[9] Even after the European Court of Justice rejected the rationale for cross border jurisdiction in patent disputes as a matter of principle on 14 July 2006, a further cross border injunction was issued by the Antwerp court of appeal on 25 April 2007 preventing a US based company and its Belgian subsidiary from offering and selling certain patented gas bottles in Europe, and allowing them only to manufacture these bottles in Belgium for export and sale outside the EP designated countries.[10]

Most likely, none of these decisions, each dealing with complex technology, would have seen the light of day if the courts had tried to examine both infringement and validity.

B.  How the “Belgian torpedo” gave a new reason to bifurcate validity and infringement

11.  Another result of the protracted nature of proceedings in the Belgian courts (in main proceedings at least) was that it attracted sham litigation, launched by alleged infringers who wished to avoid or delay foreign patent infringement proceedings against them. The Brussels court of first instance, notorious for its backlog, became the preferred venue for the so called “Belgian torpedoes”, whereby an alleged infringer sought from the Belgian court a cross border declaration of non-infringement. When this suit was brought before the patentee had himself brought a suit in another – more efficient – jurisdiction, the latter court was obliged to stay its proceedings under the international rules of lis pendens until the Belgian court had issued a final decision.

The discussion about cross border jurisdiction in patent matters has raged in Europe for more than fifteen years. The Belgian courts did not wait for the ECJ’s landmark judgements of 13 July 2006 in Gatt vs. Luk and Roche vs. Primus, but instead gave answers to these questions by distinguishing, once more, between the issues of patent validity and infringement.

12.  It is useful to remember that the English courts were the first to point out that cross border jurisdiction was excluded for them as a matter of principle because the infringement of a patent was inextricably linked with the validity of that same patent. Since the European rules of jurisdiction prevent a court from deciding the validity of a foreign patent, cross border infringement proceedings were for this reason considered impossible in England.

This reasoning was strongly reflected in the first decision issued by a Belgian court dealing with a "Belgian torpedo”. On 12 May 2000, in Röhm vs. DSM, the Brussels court of first instance relied on the inevitable interrelation between patent validity and infringement to categorically reject all types of cross border relief:

Since the European patent has to be considered, in each contracting state in which it is valid according to the European Patent Convention, as a separate national patent, the issue of (non) infringement is inherently linked with the issue of validity; As the defendants correctly argue, the sole exception to this rule arises where the plaintiffs explicitly state that they do not contest the validity of the patent, so that only infringement would be at stake in this procedure.

To the contrary however, the plaintiffs explicitly claim the invalidity of the patent – it is true, only for the Belgian territory - a claim which is however interconnected with the infringement issue; there can be no infringement if the patent is invalid; consequently, in the first place the validity of the patent has to be analyzed before the judge can analyze the issue of (non) infringement; That, indeed, an analysis of an infringement implies the existence of a valid patent; otherwise, the analysis of whether there is an infringement or not has no subject; the plaintiffs moreover would not have any standing for a judgment of infringement or non infringement if no valid patent exists. (...)

Only the judge of the state where the issue of validity is at stake has jurisdiction to rule upon infringement.(...)

The jurisdiction of this Court is thus limited to the issue of validity of the patent in Belgium and to the issue of (non) infringement in Belgium.[11]

In subsequent decisions, however, that same Brussels court took a much more nuanced position: it rejected its cross border jurisdiction for reasons that were no longer related to the link between validity and infringement. In Roche vs. Glaxo, the Brussels court of first instance held as follows:[12]

The Court finds that this [cross border] claim [for a declaration of non-infringement] is not related to the claim for revocation of the Belgian part of the European patent, but rather extends that claim beyond the Belgian borders. A decision on this claim is unrelated to the fate of the Belgian patent. The Court must, therefore, examine whether it has jurisdiction over this matter. (…)

This means that a plaintiff may not, for the sole purposes of withdrawing the (principal) defendant from his natural court, summon other defendants or a company belonging to the defendant to be summoned, so as to enable the court that has jurisdiction over the latter to also assert jurisdiction over the (principal) defendant. This would be contrary to the predictability of the court having jurisdiction, as intended by the Brussels Convention.

13.  A few months later, on 8 June 2000, although the patent in issue and the underlying facts were identical to the case which resulted in the previously mentioned decision of 12 May 2000, the same Brussels first instance court accepted that validity and infringement of a patent could be perfectly separated from each other. It accepted, therefore, the principle of granting a cross border declaration of non-infringement, provided that the general rules of international jurisdiction were respected (which would always be the case if the defendant was Belgian).

Unlike in infringement actions, a cross border declaration of non-infringement does not pre-suppose a valid patent: it is of course impossible to infringe an invalid patent, but that is not so for the opposite case: the validity of a patent, determined by a court who has exclusive jurisdiction hereto, has no impact on the question of whether a randomly selected series of facts would constitute an infringement of a valid patent. In this respect, the so-called “mirror action” for a determination of non-infringement is not inextricably linked with the question of registration or validity of a patent.

Consequently, a national Court could perfectly well decide that no cross-border infringement is or has been committed on a certain patent.  This national Court does not even have to enter the sphere of exclusive jurisdiction of his foreign colleagues regarding the question of validity of the patents filed or registered in different Contracting States for that purpose. He can solely base his reflections on the presumed validity of the national patent. So the defendants are wrong when they argue that this court should decline its jurisdiction solely on the basis of articles 16.4° in conjunction with 19 of the Brussels Convention.[13]

14.  Ultimately, in its landmark decision of 21 February 2001,[14] the Brussels appeal court confirmed the previously mentioned decision of 8 June 2000 from the lower court, and hereby sealed the fate of the “Belgian torpedo”. In doing so, however, the court did not rely on article 16.4° of the Brussels Convention. Nor did it rely on the necessity of keeping the determination of validity and infringement together. The appeal court only stressed the fact that the national patent claims in a cross border action bear no connection to each other, because after its grant a European patent is only a “bundle” of national patents whose infringement and validity are governed by national laws. Five years later, on 14 July 2006, the ECJ would confirm this approach in its landmark decision Roche vs. Primus.[15]  

C. Bifurcation in main proceedings: infringement decided first, and validity later?

15.  So far, it has been shown how bifurcation became relevant in solving problems in the framework of preliminary injunction proceedings, whether or not in a cross-border scenario. But the story does not end there. Surprisingly, bifurcation has arisen even in Belgian main proceedings, sometimes instigated by the courts, sometimes by the parties.

16.  This phenomenon can be attributed to the protracted nature of the centralized opposition proceedings of the EPO. Such drawn out proceedings put an alleged infringer at a disadvantage before the Belgian patent courts, as those courts  considered the patent prima facie valid for preliminary injunction purposes (as explained above) and refused to take a position in relation to validity during the main proceedings by systematically staying such proceedings as soon as a (counter)claim for invalidity was filed. It was argued that the fate of the European patent had first to be decided in the centralized EPO opposition proceedings. Belgian patent judges thereby gratefully deferred the complex issues of patent validity to the specialized EPO, hoping that the discussion would end or at least crystallize there, so that they would only have to deal with validity issues on the basis of an already established (but non-binding) opinion.[16]  Because of the systematic stay of the main proceedings, it took many years before an alleged infringer could obtain legal certainty about his infringement position.  In the meantime, he was exposed to preliminary injunction proceedings or even ex parte injunctions in which he could not raise invalidity arguments, because of the presumed validity of the asserted patent, and where infringement was only assessed on a prima facie basis, i.e. without any consideration as to what the judge would decide in the main proceedings. As the Belgian courts did not allow a party to seek an “anti-suit injunction” against a patentee via preliminary proceedings, the only option for an alleged infringer was to file an action for a declaration of non-infringement via regular proceedings.

As demonstrated in the previous chapter, cross border actions for a declaration of non-infringement (called “torpedoes”) were disallowed by the Belgian courts except in very narrow circumstances. However, nothing prevents an alleged infringer from seeking a declaration of non-infringement for the Belgian territory alone. So long as the alleged infringer does not assert any invalidity arguments against the threatening patent, the court is not given an excuse to stay the infringement proceedings.

17.  So this time it was not the courts, but the alleged infringers who provoked the bifurcation, hoping to speed up the main proceedings and avoid a stay. They deliberately decided to raise only non-infringement arguments and challenge the validity of the patent via separate proceedings. However, this tactic was not without risk: in two decisions from the Antwerp appeal court and the Ghent first instance court, the parties were held liable for patent infringement and subjected to permanent injunctions plus damages, notwithstanding the fact that validity proceedings against the asserted patents were still pending before the Antwerp first instance court and the EPO, respectively.[17] In both cases, the courts issued an enforceable injunction because they found that there was infringement, but did not want to wait for the outcome of the invalidity proceedings. This seemed to be justified because the alleged infringer had, after all, deliberately kept the invalidity discussion separate in order to avoid a stay. Much less understandable was that these decisions also stated that, even where the patent was later revoked in the separate validity proceedings, the patentee would not have to compensate the infringer for damage eventually suffered by him due to the enforcement of the injunction. This reasoning was supported by a peculiar reading of article 50.2 ° of the Belgian Patent Act, which provides that the nullity of a patent has no retro-active effect vis à vis judgements containing an injunction or an award of damages that have become final prior to the patent’s revocation.

With such an unbalanced reasoning, not only did the infringement discussion become separated from the validity discussion (by a voluntary choice of the defendant) but the courts took the bifurcation even further by severing the patentee’s liability for wrongful enforcement of his patent from the question of patent validity. This gave patentees substantial immunity from liability suits.

The Belgian Supreme Court never had the chance to take a position on this because, as will be shown below, further changes in the Belgian procedure have substantially accelerated patent litigation, such that the problem described above can no longer arise. However, ironically, another sort of bifurcation has now evolved because the proceedings are (too) expeditious.

D. The “fast track action”: the final route towards bifurcation?

18.  Due to changes in the substantive and procedural patent enforcement laws, patents can now be asserted in Belgium via the so-called “fast track action” route. This procedure is well known and has been widely used in Belgium since the 1970’s for claims against acts of unfair competition. It was available to owners of copyright and trade marks since 1996 and 2004, respectively, and since November 2007 patent owners have also been able to rely on this procedure. A patent “fast track action” is heard by the President of one of the five Commercial Courts (rather than the slower first instance courts) within a matter of months. The court sits with a single judge, who must assess the case as if it were proceedings on the merits.  A plaintiff can only obtain an injunction, not damages or other sorts of relief, but the President’s opinion in relation to validity and infringement will bind the “regular” courts, who would afterwards be asked to rule about damages or other aspects of the case. The legislator explicitly provided for the possibility to bring a counterclaim for invalidity, and have the asserted patent revoked in such proceedings.

19.  For reasons exactly opposite to those in the scenarios described above, the “fast track” judges also decided to bifurcate the issues of validity and infringement. It was ruled that since the goal of these proceedings focused on issuing “fast” injunctions, and the acting judges sat alone (rather than in a panel of three judges as for “regular” proceedings), it was not within their remit to systematically evaluate patent validity, and potentially revoke them, in the absence of any clear reason to do so. It was decided that patent validity should only be examined after it had been first determined that there was patent infringement. This reasoning was developed in the first “fast track action” decision rendered by the President of the Brussels Commercial Court[18] and was re-iterated in a decision of the President of the Ghent Commercial Court in August 2009.[19]

So whereas in “regular” proceedings a Belgian court would typically look first at patent validity (and go no further if the patent were found invalid) before considering the infringement issue, the “fast track” judge will address the issues the other way around: he will start with an assessment of patent infringement, and only after finding infringement, will he examine the invalidity attack and declare the patent valid, revoke it or amend it. If he finds there has been no infringement, he will decline jurisdiction in relation to the invalidity counterclaim. This route offers an important strategic advantage for a patentee because he will only expose his patent to an invalidity challenge if he has already won on infringement. This also comes very close to a truly bifurcated system. It remains to be seen whether the Belgian appeal courts will approve this reasoning, which was not perhaps intended by the legislator, but has found support with commentators and practitioners.

E. Confidential information in a report of “saisie description”: another reason to apply bifurcation?

20.  In the past year and a half, there has been a further remarkable development, in which the Belgian courts found yet another reason for imposing a different treatment to infringement issues vis à vis validity issues. It occurred in the framework of discussions on how to protect sensitive information about an allegedly infringing product, disclosed via a Belgian report of “saisie description”, when there was a risk that the patentee would amend his patent on the basis of that information.

As described above, the procedure of “saisie description” enables a patentee to gain access to a competitor’s product, the details of its manufacturing process and even the composition of the allegedly infringing product. It is widely accepted that the proprietary or sensitive nature of such business secrets is no impediment for disclosing this information to the patentee via the court expert’s report, so that the patentee can assess whether his patent is infringed or not. The question was to what extent the patentee can use this information during the subsequent main proceedings, where he is likely to face a challenge against the validity of his patent. Should he use it only to assess the likely outcome of the infringement action, or can he also use it to assess the extent to which the claims of his patent could be amended/limited without loosing the benefit of his infringement argument?

Two decisions from the Brussels appeal court[20] and the President of the Brussels commercial court[21] have taken the view that a patentee should not rely on information disclosed by the alleged infringer during a “saisie description” for the purpose of making amendments to his patent during (or to avoid) opposition proceedings before the EPO. In the first decision, rendered during the main proceedings, the appeal court held this to be a violation of a confidentiality undertaking and declared that the auxiliary claims of the patent were inadmissible. In the second case, proceedings were brought by the alleged infringer to oppose an ex parte order for “saisie description” so that the information obtained via the court expert could not be shared with the patent attorneys in charge of defending the asserted patent before the EPO. In this case, to avoid abuse of the report’s information a separation of information was ordered, relying on section 1369 bis / 7.2° of the Belgian Judicial Code, which states that the report resulting from the “saisie description” shall be treated as confidential and can only be used in Belgian or foreign main proceedings or preliminary injunction proceedings. In both decisions, the courts considered that the discussions as to the validity and infringement of a patent could be perfectly well separated from each other, leaving no reason to share information about the allegedly infringing product with those preparing a defence against the invalidity attack. In a court system where issues of validity and infringement ought to be considered simultaneously and are intrinsically linked, such reasoning is astonishing.


21.  Although Belgian patent litigation professes to be a “combined” system, the past ten years have seen many instances where the questions of validity and infringement have been separated from each other, either by the Belgian courts or by the parties. In each case, this was inspired by a concern for rendering a more efficient procedure or to preserve confidentiality. Whether it was fair in each case, or legally justified, is debatable. What is clear is that without these repeated instances of bifurcation, the Belgian patent courts would not have made such substantial progress in the speed and quality of their decisions as has been witnessed in the past decade. These examples demonstrate how in jurisdictions where patent litigation is perceived as complex and time consuming, bifurcation can offer a solution.

[1] As per Professor Mario Franzosi, as referred to by  Jacob LJ in European Central Bank v Document Security Systems Inc [2008] EWCA Civ 192, para 5.

[2] European Central Bank v Document Security Systems Inc [2007] EWHC 600 (Ch), per Kitchin J at para 88.

[3] For an example see Ghent court of first instance, 12 June 2002 (N.V. Roxell vs. SKA S.p.a.) IRDI 2002, 122, with comment by R. Peeters and B. Vandermeulen.

[4] See article 41, 4 juncto article 42, 1 of the current draft agreement of 14 December 2005 of the working party on litigation regarding the establishment of a European patent litigation system  

[5] These were the legal criteria set forth in the Belgian Supreme Court’s leading decision of 16 November 1984, Ing. Cons. 1985, 164, with comment by F. de Visscher.

[6] Antwerp court of appeal, 25 June 1990 (Improver vs. Remington) Ing. Cons. 1991, 249; Brussels court of Appeal, 14 June 2004 (Synthon vs. SKB) IRDI, 2005, 67; Antwerp court of appeal, 29 March 2003 (Fort Koffiebranderij vs. Sara Lee), IRDI 2003, 256; Antwerp court of appeal, 27 April 2007 (Praxair vs. ATMI) IRDI 2007, 248, with comment by Kristof Roox.

[7] Antwerp court of appeal, 18 October 1999, IRDI. 2002, 292.

[8] Pres. Brussels court of first instance, 14 September 2001 (Colgate vs. Unilever) IRDI 2002, 239.

[9] Pres. Brussels court of first instance, 25 March 2005 (Altana Pharma), IRDI 2005, 302, with comment by Kristof Roox.

[10] Antwerp court of appeal, 25 April 2007, Praxair vs. ATMI, IRDI 2007, 248, with comment by Kristof Roox.

[11] Brussels court of first instance, 12 May 2000 (Rohm Enzyme vs. DSM), IRDI 2002, 321.

[12] Brussels court of first instance, 8 June 2000 (Roche vs. Glaxo), IRDI 2002, 313.

[13] Brussels court of first instance, 19 January 2001  (Novo Nordisk vs. DSM), IRDI 2002, 304.

[14] Brussels court of appeal, 20 February 2001 (Roche vs. Glaxo) IRDI 2002, 284, with comment by Steven Cattoor.

[15] Roche vs. Primus, ECJ 13 July 2006, C-539/3 par. 30 et seq.: “It is apparent from Article 64(3) of the Munich Convention that any action for infringement of a European patent must be examined in the light of the relevant national law in force in each of the States for which it has been granted. It follows that, where infringement proceedings are brought before a number of courts in different Contracting States in respect of a European patent granted in each of those States, against defendants domiciled in those States in respect of acts allegedly committed in their territory, any divergences between the decisions given by the courts concerned would not arise in the context of the same legal situation. Any diverging decisions could not, therefore, be treated as contradictory. In those circumstances […], it is clear that such a connection could not be established between actions for infringement of the same European patent where each action was brought against a company established in a different Contracting State in respect of acts which it had committed in that State.”

[16] See article 138 of the European Patent Convention (EPC).

[17] Antwerp court of appeal, 8 November 2005 (Sara Lee vs. Koffiebranderij Fort, Beyers and Cafés Liegeois), IRDI 2006, 134; Ghent court of first instance, 23 May 2005 (Logitec vs. Constructie Werkhuizen Demaitere) IRDI 2006, 169.

[18] Pres. Brussels Commercial Court, 9 October 2008, no appeal (MaxCard vs. Master Card) IRDI 2009, 129, with a positive comment by Prof. Dr. Andrée Puttemans.

[19] Pres. Ghent Commercial Court, (Machinefabrieken Bollegraaf & Grumbach vs. Wagenseld B.V. & Stora Enso N.V.), unpublished, currently on appeal.

[20] Brussels Appeal Court, 20 June 2008 (GSK Biologicals vs. Sanofi Pasteur) IRDI 2008, 345.

[21] Pres. Commercial Court of Brussels, 7 May 2009 (GSK Biologicals vs. Novartis) IRDI 2009, IRDI 2009, 355, with critical observations by Bruno Vandermeulen.

First published in "Festschrift für Wolfgang von Meibom zum 65 Geburtstage", Carl Heymans Verlag, 2010, p. 469