Glaxo Wellcome UK Ltd (t/a Allen & Hanburys) & Anor v Sandoz Ltd & Ors

The claimants applied by application notice dated 22 August 2018 for an order permitting the second claimant to use certain documents disclosed by the 1^st to 4^th defendants (“the Sandoz Defendants”) in this claim in a claim in Belgium between the second claimant and Sandoz NV (“Sandoz Belgium”). The two claims are part of global litigation between members of the GlaxoSmithKline and Sandoz groups of companies. In Europe there are claims in several jurisdictions including England and Wales, The Republic of Ireland, Germany, The Netherlands and Belgium.

Seretide (known as Viani in Germany and Advair in the USA), which is now out of patent, is used by patients for the treatment of asthma and chronic obstructive pulmonary disease (COPD). It is a combination of two preventers, a Long-Acting Beta-Agonist and an Inhaled Corticosteroid. The claimants say it was an innovative product at its launch. It is sold in two types of inhaler; a ‘boot shaped’ metered dose inhaler (MDI) and a dry powder disk inhaler (DPI). The MDI is marketed in the UK under the name ‘Evohaler’ and the DPI under the name Accuhaler. Both are available in different dosages.

The second claimant was the registered proprietor of a colour Community Trade Mark number 3890126 for the combination of purple 2587C and purple 2576C registered with effect from 16 June 2004. However, the Community Trade Mark was declared invalid on 11 November 2016 on the counterclaim by the Sandoz Defendants in these proceedings by HH Judge Hacon (sitting as a judge of the High Court) and his decision was upheld by the Court of Appeal on 6 April 2017.

These proceedings now comprise solely a claim by the claimants in passing off. The core claim is set out in paragraphs 13 and 14 of the re-re-amended particulars of claim:

“13.

The Claimants’ Seretide Accuhaler and Evohaler products are sold in a getup and packaging which are distinctive of the Claimants. The Claimants or one or more of them are the owners of a significant goodwill in the business of selling Seretide Accuhaler and Evohaler products in the UK which has become associated in the minds of the trade and public with the particular get-ups and packaging in which the said products are sold.

14.

In this action the Claimants will rely on the following distinctive indicia either alone or in combination (hereinafter ‘the indicia’):

14.1

as regards the get-up of the Seretide Accuhaler inhaler itself (as shown in the examples below):

14.1.1

the use of the purple colours (each either alone or in combination as independent distinctive indicia);

14.1.2

the respective positions of the purple colours around the product and the central position of a white label; and

14.1.3

the overall shape and size of the Seretide Accuhaler product itselfin the open and/or closed

position;

14.2

the use of the colour purple and white on Seretide Accuhaler packaging as shown in the examples below;

The Sandoz Defendants launched a generic product named “AirFluSal Forspiro” in the United Kingdom in 2015 and elsewhere internationally variously from 2014 onwards. The products are both prescription-only and they are marketed and sold as boxed inhalers.

Pictures of the AirFluSal product as it was launched in Ireland are below:

The packaging of the two competing products can be seen from the following images:

The claimants allege that the Sandoz Defendants chose AirFluSal Forspiro’s get up with the deliberate aim of deceiving or creating confusion in the mind of the relevant public. The Sandoz Defendants accept that the question of whether the public are deceptively confused may take into account the Sandoz Defendants’ intentions and documents relating to the design history of the AirFluSal Forspiro dating back to 2003 have been disclosed by the Sandoz Defendants in this claim.

The trial of this claim was due to take place in October 2018. However, on 3 July 2018 Mr Rosen QC (sitting as a deputy judge of the High Court) made an order vacating the trial date and directing that the fifth, sixth and seventh defendants be joined as parties to the claim. The trial has now been re-fixed to take place in July 2019. The fifth defendant, Sandoz AG, is part of the Sandoz group of companies. The sixth and seventh defendants are independent companies (together “Vectura”) which were involved in the design of the AirFluSal Forspiro product.

The disclosure exercise between the claimants and the Sandoz Defendants has been very substantial. It involved the Sandoz Defendants reviewing 406,300 documents using 50 legally qualified reviewers. This led to the subsequent disclosure of slightly in excess of 75,000 documents to the claimants.

There is a marked contrast in the manner in which litigation is conducted in England and Wales on the one hand and Belgium (and most other Civil law countries) on the other hand. In England and Wales, the ability to obtain disclosure that is adverse to the other party’s claim is an important feature of litigation. However, the evidence provided in connection with the application shows that disclosure is only available in a very limited form in Belgium. One of the issues to be determined is whether disclosure obtained in this jurisdiction should be made available to a party that is engaged in litigation in a jurisdiction where disclosure, if not unknown, is very limited in scope.

The claimants’ application is to use the documents in the existing proceedings in Belgium and any new action the second claimant brings in Belgium against to second to seventh defendants. What the claimants have in mind is explained in more detail in the 10^th witness statement of Eifion Morris (“Morris 10”). The uses to which the second claimant wishes to put the documents are:

“(i)

to seek advice from its Belgian counsel as to whether the documents merit being adduced as evidence in the Belgian action on the basis that they correct factual misstatements and disprove arguments made by Sandoz Belgium in those proceedings;

(ii)

to seek advice from its Belgian counsel as to whether the documents merit bringing proceedings in Belgium against the second, third, fourth, fifth, sixth and/or seventh defendants in Belgium; and

(iii)

if the second Claimant’s Belgian counsel believes that they merit the action described above (in whole or part) then to use the documents for those purposes in the Belgian action and any new action against the second, third, fourth, fifth, sixth and seventh defendants in Belgium”

The application was dealt with on the basis that if an order is to be made, use of the documents is contingent upon the second claimant first obtaining advice from its lawyers in Belgium about the appropriateness of using them, or some of them, for the purposes Mr Morris explains. But it is not part of the second claimant’s application

EU Directive 2008/95/EC (as recast in EU Directive 2015/2436) “The Trade Mark Directive” has been implemented in the Benelux Convention on Intellectual Property. On 2 July 2015, the second claimant became the registered owner of a Benelux trade mark registered under number 0977861 used in relation to Seretide. The registration is for a single colour described as “Violet 2587C” in classes 5 and 10 (“the Benelux mono-colour trade mark”).

The trade mark registration covers Belgium, the Netherlands and Luxembourg. The registration pre-dated the launch of the AirFluSal Forspiro product in Belgium and the launch has not yet taken place.

On 3 November 2015, Sandoz Belgium, which is the holder of marketing authorisation for the AirFluSal Forspiro product in Belgium, brought a claim before the Dutch-speaking Court of Commerce of Brussels (“the Belgian Court”) seeking, amongst other things, a declaration of invalidity in relation to the Benelux single colour registration (“the Belgian Claim”).

There has been some issue between the parties about the scope of the Belgian Claim. Mr Christian Dekoninck, who represents Sandoz Belgium in the Belgian claim, describes the issues in the following way:

“1.1

The Benelux mono-colour trademark cannot be considered distinctive ab initio. European and Belgian case law confirms with regards to single colours that distinctiveness without the prior use is inconceivable save in exceptional circumstances. No such circumstances were established by the second claimant.

1.2

Second, the Benelux mono-colour trademark is merely descriptive as it only serves to designate the purpose (combination product), composition or the strength of the medicinal product.

1.3

Third, the second claimant does not establish that the colour trademark may have acquired distinctiveness through use.

1.4

Finally, the registration of the Benelux mono-colour trade mark is contrary to the public interest as it unduly limits the possibility of other manufacturers to use the basic colour purple for descriptive purposes.”

According to the second claimant, Sandoz Belgium has argued in the Belgian Claim that:

The second claimant has defended the Belgian Claim and counterclaimed for infringement of the Benelux trademark. Although (surprisingly) it was not mentioned by Mr Dekoninck, it is now accepted that the second claimant’s counterclaim includes a claim under article 2.20(1)(c) of the Benelux Convention on Intellectual Property (reflecting the Trade Mark Directive Article 10(2)(c)) for taking unfair advantage. The unfair advantage claim has been a subsidiary part of the second claimant’s counterclaim in Belgium, but it has assumed significance in the application because the type of documents that may be relevant to it are wider than in relation to the other issues.

On 12 January 2017 the Belgian Court decided that the Benelux mono-colour trade mark was valid but was not infringed by the possible marketing of the AirFluSal Forspiro inhaler. Both parties have appealed that decision and both parties have already filed appeal briefs in accordance with a trial schedule agreed upon between them. It is now common ground that the appeal is a re-hearing of the claim. At one time it was said the second claimant’s application was urgent and the Sandoz Defendants questioned whether further evidence would be admitted by the Belgian Court. It is now accepted that the court in Brussels will abide the outcome of the application before me and it is open to both parties to file additional evidence with the second round of appeal briefs. The parties have agreed that, if the second claimant is successful on its application, to a greater or lesser degree, sequential additional briefs will be filed with the second claimant going first. This will be followed by an oral hearing of the appeal.

CPR 31.22 provides:

“(1)

A party to whom a document has been disclosed may use the document only for the purpose of the proceedings in which it is disclosed, except where –

(a)

the document has been read to or by the court, referred to, at a hearing which has been held in public;

(b)

the court gives permission; or

(c)

the party who disclosed the document and the person to whom the document belongs agree.”

The principles that apply to an application made to obtain the court’s permission to use documents for a purpose other than the proceedings in which the documents were disclosed has been considered in a number of authorities. Mr Hickman who appeared for the claimants provided a helpful summary of the authorities which I adopt.

Mr Hickman also drew attention to the decision in Cobra Gold Inc v Rata [1996] FSR 819 in which Laddie J set out a number of principles he considered the court should take into account. His observations were referred to as being ‘useful’ in SmithKline. The court should consider:

There are additional considerations which need to be taken into account where the documents are proposed to be used in foreign proceedings:

I will consider in due course how these principles apply to this application. It is worth observing, at this stage, that the particular aspects of the application in this case that merit noting are:

Before setting out the issues between the parties in the Belgian Claim and considering the degree of overlap between the two claims, it is necessary to make some introductory remarks about how this application was dealt with in view of the fact that the request is for permission to use multiple documents. The initial application was for permission to use over 100 documents. On the first day of the hearing, I expressed the view that the court would need to review each of the documents for their relevance to the Belgian Claim, unless they could be put in classes and samples in each class reviewed. The hearing was adjourned part heard to enable the claimants to consider their position. The outcome of the review was that although Mr Hickman maintained his submission that the court was not required to consider each document, two changes were made to the application. First, the number of documents for which permission was sought was reduced to 51. The claimants produced a set of those documents with redactions to remove irrelevant material on 2 November 2018 in two document bundles. Secondly, the claimants provided a document that I will describe as a ‘Relevance Schedule’. It described each of the 51 documents and provided reasons in summary form why each was needed in the Belgian Claim.

For my part, I do not see how the court can deal with an application such as this one without reviewing each document, unless the parties have agreed upon an alternative approach, such as reviewing a sample. When reviewing an individual document, the court will, of course bear in mind that individual documents may have greater relevance or weight (or both) when seen in the light of other documents.

The sampling approach is more obviously suitable where the documents are an homogenous class. However, where, as here, the documents are sought to be used for one of two distinct uses and there are nine different reasons for relevance put forward by the claimants (with each document said to be relevant for one or more different reasons), individual review is the only option. This is potentially very time consuming and it was helpful that the claimants halved the number of documents the court had to consider.

It seems to me that where the court is asked for permission to use documents in foreign proceedings, it is not necessary for the court to do more than consider whether the documents are likely to be of relevance to the foreign proceedings. It is not for the court to determine whether the documents will, in fact, be thought to be sufficiently important to warrant introduction in the foreign claim by the foreign lawyers who review them, or that the foreign court will find them to be sufficiently compelling to give them weight in that court’s determination. This is so for largely pragmatic reasons. ‘Likely relevance’ is an initial threshold following which the court will wish to consider the broader issues that are to be derived from the authorities, namely whether there are special circumstances that provide cogent reasons why permission should be given.

The Belgian Claim has been summarised in the evidence of Mr Dekoninck to which I have already made reference, Mr Bruno Vandermeulen of Bird & Bird LLP in Brussels and Ms Florence Verhoestrate a partner with NautaDutilh BVBA/SPRL in

Mr Dekoninck’s witness statement focuses on issues concerning the validity of the mono-colour trade mark and infringement and contends that the intention of the parties “is in general irrelevant”. It was left to Ms Verhoestraete to point out that the counterclaim includes a claim for unfair advantage. Such a claim requires the court to have regard to whether there was an intention to ‘ride on the coat tails’ of the mark

Mr Dekoninck makes two further points that warrant mentioning. He says:

There was at one time an issue between the parties about the maintenance of confidentiality if documents disclosed in these proceedings were permitted to be used in the Belgian Claim. It is now common ground that the court file in Belgium is not open to the public and that is open to the parties to agree a contractual confidentiality regime. The second claimant has offered to enter into such a regime and it seems to me it cannot seriously be contended that a contractual regime, albeit one that it is not subject to the court supervision and the ultimate sanction of contempt, is materially less satisfactory than the type of regime which would be applicable in this jurisdiction.

Evidence in support of the application has been provided in the 10th and 16th witness statements of Mr Eifion Morris, a partner in Stephenson Harwood LLP, on behalf of the claimants and the 10^th witness statement of Mr Marcus Collins, a senior associate at White & Case LLP, on behalf of the Sandoz Defendants. I have had regard to their evidence, but it is unnecessary to refer to it in detail.

I shall now explain how the Relevance Schedule is set out. It is divided into six components:

Open documents are not subject to the agreed confidentiality club or pro tem confidentiality orders. Closed documents are subject to that regime.

Morris 7 was made in connection with the claimants’ application to join the 5^th to 7^th defendants; the application that was heard by Mr Rosen QC on 3 July 2018. It referred to and exhibited a number of documents in making out a case that the court should make an order for joinder.

Morris 9 was made in support of an application to use documents disclosed in these proceedings for the purposes of related proceedings in the Republic of Ireland. They comprised documents forming exhibits to Morris 7 and additional documents. The claim in Ireland is, like this claim, based on the tort of passing off and the application to use the documents was agreed. The Sandoz Defendants say that it is quite different, on the one hand, to use disclosure documents where the cause of action is almost precisely the same to, on the other hand, use documents where the cause of action and the approach of the court toward the use of disclosure documents is quite different. In any event, it seems to me the fact that the Sandoz Defendants’ agreement to documents being used in Ireland carries little or no weight when the court considers the application of the relevant principles in this application.

As I have indicated Morris 10 is the statement made in support of this application.

Each of the documents listed in the relevance schedule is given a unique exhibit number and the Bates number allocated to the document in the Sandoz Defendants’ disclosure is also provided. For convenience and clarity, I will refer to the exhibit number when referring to an individual document.

There are nine relevance categories. The order in which they are listed is not indicative of the importance or weight that attaches to each category.

It is necessary to consider each of these points, although there is some overlap between them. The correct approach, in my judgment, is to form a view at a relatively high level about whether each of these issues is material to the Belgian Claim. If in any case the issue does not arise in the Belgian Claim, or is peripheral to it, it should be disregarded. If, however, it is clear that the issue is at least arguably engaged in the Belgian Claim, it is then necessary to consider whether any of the documents are likely to be relevant to the Belgian Claim in response to one or more of the relevance issues.

Sandoz Belgium contends in the Belgian Claim that the colour purple reflects the active pharmaceutical ingredients and an industry colour convention; this can be seen from paragraphs 17 to 21, and 90 to 98 (and footnote 74) of Sandoz’ Appeal Brief. It is said the choice of the colour purple was dictated by these considerations. At paragraph 21 of the Brief, as part of the Conclusion, it is stated that:

“It is clear from the above that colours used on inhalers for the treatment of asthma and COPD are mainly descriptive because they have a functional role in identifying or distinguishing the active pharmaceutical ingredient and the strength thereof. These colours indeed indicate the nature and strength of the product. Patients and health professionals rely on these colours when using or explaining the use of such inhalers. The importance of a colour-coded asthma treatment in patient education is indeed internationally well accepted…”.

The claimants say that when the colour purple was chosen for the Sandoz product, Sandoz had global colour rules and had they been applied the active ingredients in the AirFluSal Forspiro would have been shown using blue. It is clear that the issue about an alleged deviation from Sandoz’ own colour convention is equally relevant to packaging as it is to the device. Thus, documents said to be relevant to Sandoz’ colour code for packaging are potentially relevant to the Belgian Claim.

The Sandoz Defendants deny that the colour purple is distinctive of Seretide or that it has brand identity with Seretide.

Mr Howe refers to the decision in Oberbank AG v Deutsher Sparkassen und Giroverband eV Joined Cases C-217/13, C-218/13 ECLI: EU:C: 2014:2014 and points out it is necessary to show not just brand association with a product but also that the colour is a badge of trade origin. Since the claimants’ product was the only one on the market for many years it is unsurprising that surveys contain spontaneous references to Seretide. I consider that the studies resulting from surveys and focus groups the second claimant wishes to rely on are likely to be treated with caution in the Belgian Claim. The findings on a survey that does not include any of the Benelux countries or is from an earlier period may not be regarded as providing assistance. That said, Sandoz Belgium’s Appeal Brief refers to proceedings in other countries and

Documents are said to show that purple is used by Sandoz as a trade mark and as a key component in its branding. Paragraph 124 of the Sandoz Belgium Appeal Brief asserts that Sandoz does not use the colour purple to distinguish the AirFluSal Forspiro product. It goes on:

“The use of another shade of purple, in combination with white on the inhaler and its packaging by Sandoz is indeed only an indication of the characteristics of the product. This indication refers to the fact that [the product] is a combination product, in accordance with the informal colour practice for medicinal products for the treatment of COPD/Asthma.”

Documents that may show Sandoz used the colour purple as a trade mark, or which might tend to contradict or cast doubt on the assertion I have quoted from the Appeal Brief, or other similar statements, are likely to be relevant to the Belgian Claim.

The extent to which Novartis does not conform to colour codes is said to be in issue.

I have already referred to the decision in L’Oreal v Bellure and the possibility that documents showing how Sandoz went about the production of a generic product, and its choice of colour, could be relevant. It is right that the unfair advantage claim by the second claimant is subsidiary to the main thrust of its counterclaim. That however is not a relevant consideration when considering likely relevance for the purposes of this application. It may that be that in light of consideration of the documents the second claimant wishes to use, the unfair advantage claim will assume a greater prominence.

Sandoz Belgium sets out submissions at paragraphs [110] to [114] in its Appeal Brief under the heading: “The colour trade mark of Glaxo is contrary to the public interest in view of the specific circumstances of the case”. The case is developed at paragraph [111] and submissions are made about the limited availability of the basic colours for asthma and COPD inhalers “… in view of the fact that several colours are already used by many actors (only 12 basic colours can be distinguished and less than 5 would remain free) and, on the other hand, to the descriptive character of colours, which therefore must remain freely available to anyone. The availability of these descriptive signs should not be unduly limited for other market players offering the same types of goods …”. There is a public policy issue which is live in the Appeal and to the extent there are documents which contradict or weaken the case put forward by Sandoz Belgium, they are likely to be relevant.

The case put forward by Sandoz Belgium in its Appeal brief at paragraph [154] is to the effect that if the second claimant is able to enforce the mono-colour Benelux mark, it would be abusing its dominant position and Sandoz Belgium would have to redesign the AirFluSal Forspiro product. “This would result in the need to redevelop and remanufacture the new products and to obtain new marketing authorisation.” Documents showing how difficult or easy it is to re-manufacture the product have potential relevance. The position is not as Mr Howe submitted that obtaining new marketing authorisation is the only factor that may cause delay in gaining entry to the market.

The claimants say they wish to consider bringing a claim against new parties in Belgium. It is suggested that there could be no entitlement to do so. Mr Howe submits that (a) the involvement of other parties in the development of the Sandoz product is well known and the second claimant does not need the documents upon which it wishes to rely to obtain advice and bring a claim and (b) in reality, from a commercial perspective, there is little reason why the claimants might wish to do so. As to (a), it is not for the Sandoz Defendants to decide which documents the second defendant may need. That is a matter for the second defendant’s lawyers. As to (b), similarly, the second claimant is entitled to seek advice about steps it may wish to take and, if there are documents that are likely to be relevant to that issue, they should not be ruled out at the initial relevancy stage. Documents such as the contractual arrangements between Sandoz and Vectura, from which the financial arrangements have already been redacted, are plainly relevant to issue of Vectura’s involvement.

As I have indicated, the relevance criteria overlap and most of the documents which the second claimant wishes to use are said to be relevant by reference to more than one of the relevance criteria. My conclusion is that with the exception of issue (5) all the issues are helpful markers of relevance.

It is not proportionate to give reasons for accepting relevance for each document individually. My conclusion, having reviewed the Relevance Schedule and documents themselves, is that all the documents the second claimant wishes to use are likely to be relevant to the Belgian Claim apart from documents with exhibit numbers 58, 60, 99, 152, 162, 164, 165, 169.

The application is limited to a request for permission to use the documents applying the power under CPR 31.22(1)(b). There is no application for a declaration that particular documents have been “read to or by the court, or referred to, at a hearing which has been held in public”. Mr Hickman submits that the use of the documents exhibited to Morris 7 in open court is a powerful factor in relation to an application for permission to use them. He does not, however, seek to argue that an exhibit is for these purposes part of the witness statement to the effect that by reading the witness statement the court is deemed to have read the exhibits.

The documents exhibited to Morris 7 were provided to the court in order to make the case that the Vectura defendants were involved. It could not be suggested that their use was in any way gratuitous. The authorities concerning the use of documents in court have recently been considered by the Court of Appeal in Cape Intermediate Holdings Ltd v Dring [2018] EWCA Civ 1795. I need only refer to paragraphs [103] to [109]. Documents that the court was asked to read outside court will be deemed to have been read unless it can be established to the contrary. It is clear from the skeleton arguments that were filed in advance of the hearing before Mr Rosen QC that he was asked to read the witness statements, but not the exhibits, and it can be presumed that he did that. The text of Morris 7 itself refers extensively to the documents that formed the exhibits, sometimes citing extracts from them. It follows that although the exhibits themselves will not be treated as being read out in in open court, they have been extensively considered at a hearing in open court. A limited number of documents were subject the confidentiality club order and a pro tem confidentiality order that covered the hearing, but the majority were not confidential. I agree with Mr Hickman that use of the Morris 7 documents for the joinder application is a factor the court should take into account on an application under CPR 31.22(1)(b) whilst being careful not to conflate the test under 31.22(1)(a) with the test under 31.22(1)(b).

Similar considerations do not apply to the documents that are exhibited to Morris 9 or

The principal additional submissions Mr Hickman puts forward can be summarised briefly:

Mr Howe submits that the English court should be slow to permit a party to use documents obtained in proceedings in England under our disclosure procedures in a foreign claim where disclosure of a company’s internal documents will only very rarely be permitted. I can see the force of this submission if it can be suggested that a claim was brought in England for the principal purpose of obtaining disclosure and using it elsewhere. However, that could not be said of this claim. At the time it was was commenced, the claimants were seeking to enforce their Community Trade Mark within this territory. The proceedings now proceed relying on the domestic remedy of passing off. To my mind, the limited nature of disclosure in the Belgian Court is a factor of limited weight. It has the effect that an English court might be more circumspect about the Belgian Claim being flooded with a large volume of documents. The application is now limited to a relatively small number of documents and the basis for the use of each document has been carefully explained.

There is more force in Mr Howe’s submission that the court should be concerned about the claimants ‘cherry picking’ documents from disclosure in this claim and using them out of context. This might have the effect of forcing Sandoz Belgium to review a large number of documents and being required to submit a disproportionate volume of documents to correct false impressions created by disclosure documents that have been taken out of context. Mr Howe also submits that the difference of approach to the conduct of claims between England and Belgium is a material factor. He points out that in England, documents are often explained and given life in the course of cross-examination. Documents which tend to point in one direction can be explained and put in a proper context. In Belgium, although the hearing of the appeal will involve oral submissions based upon the written briefs, there will be no witness evidence to provide necessary illumination to the documents.

The court is required to balance the competing interests of justice. On the one hand there is an interest in permitting the second claimant to use documents in the Belgian Claim and to consider bringing fresh proceedings in Belgium and on the other hand in preventing injustice to the Sandoz Defendants and Sandoz Belgium. The claimants must be able to satisfy the court there are special circumstances which constitute special reasons for permitting collateral use. In my judgment, the claimants are well able to meet this test, with the corollary that the balance of justice comes down firmly in their favour. My reasons are:

I will make an order permitting the second claimant to use the redacted documents included in the two document bundles that were produced on 2 November 2018, other than those that I have excluded.