Teva Pharma - Produtos Farmaceuticos Lda & Anor v Astrazeneca-Produtos Farmaceuticos Lad & Anr

The appellants (to whom I will refer as “Teva”) appeal, with the permission of Leggatt J, his Order dated 30 June 2017 whereby he dismissed their application for summary judgment and ordered them to pay the costs of the respondents (to whom I will refer as “AstraZeneca”). The purpose of the hearing before the judge had been (pursuant to the earlier Order of Blair J dated 9 June 2017) to determine whether Teva were entitled to a declaration that, pursuant to the terms of a Settlement Agreement dated 25 February 2013, Teva and their affiliates were entitled to use, import, store, offer for sale, sell, market and distribute the Teva Product as defined in Portugal from 3 July 2017 onwards. The hearing before the judge (and the appeal before this Court) concerned a short point of construction of that Settlement Agreement.

The appellants are Portuguese pharmaceutical companies in the Teva Group. The first respondent is a Portuguese pharmaceutical company in the AstraZeneca Group. The second respondent is a Japanese company and registered proprietor of the patent EP 0521471 (“the Patent”) over rosuvastatin (“the patented compound”), a chemical compound that is the active ingredient in certain statins drugs. The first respondent held an exclusive sub-licence to exploit the patent in Portugal and was the marketing authorisation holder for drugs containing the patented compound as an active ingredient, sold on the Portuguese market.

European patents granted under the European Patent Convention have a life of 20 years from filing of the application. As the application for this patent was made on 30 June 1992, the Patent expired on 30 June 2012. Since the pre-clinical testing and clinical trials of pharmaceutical products usually continue long after an application for a patent has been made, it can take a while before the product is authorised to enter the market, by which time some of the twenty year period will have expired. To address this problem, Supplementary Protection Certificates (“SPCs”) were created by (most recently) Regulation (EC) No 469/2009 of the European Parliament and of the Council.

AstraZeneca’s SPC in this case was granted on 25 November 2003 and lasted until 3 July 2017. The expiry date was subsequently corrected to 30 June 2017 by an ex officio decision of the Portuguese Patent Office on 1 July 2013, but at the time that the Settlement Agreement was concluded, the expiry date of the SPC was 3 July 2017.

In Europe, generic pharmaceutical manufacturers such as Teva are entitled to obtain marketing authorisations prior to the expiry of a patent or SPC and it will be for the patentee to enforce its intellectual property rights before the national courts or, in the case of Portugal, before mandatory arbitration tribunals.

In January 2012, Teva had obtained marketing authorisations from the relevant Portuguese authority, Infarmed, for their generic rosuvastatin product. The response of AstraZeneca was to commence proceedings in the Portuguese Administrative Court against Infarmed, in which Teva were interested parties, and mandatory arbitration proceedings against Teva in which the tribunal was established on 31 July 2012. In those arbitration proceedings, Teva served a lengthy Defence on 17 November 2012, in which, inter alia, they challenged the validity of the Patent. On 2 December 2012, AstraZeneca served a detailed Reply, which, inter alia, contended that the arbitrators were not a forum before which the validity of the Patent could be challenged.

By the end of January 2013, AstraZeneca were concerned that a launch by Teva of their generic product was imminent, so they sought a preliminary injunction in Portugal to protect their rights under the SPC. In fact, Teva launched their product in Portugal on 1 February 2013. Following correspondence between the parties, the Settlement Agreement was entered into to settle disputes relating to the sale of the generic product.

The Preamble to the Settlement Agreement identified the parties to the Agreement and stated at (v) that AstraZeneca had commenced “the Proceedings” defined in clause 1.1 as encompassing the proceedings before the Portuguese Administrative Court, the arbitration proceedings and various injunction proceedings. Recitals (vi) and (vii) in the Preamble stated:

Clause 1 contained a number of Definitions. So far as relevant, “Patent” was defined as “EP 0521471 and Portuguese Supplementary Protection Certificate 156”. “Product” was defined as AstraZeneca’s product containing the patented compound. “Generic Product” was defined as any pharmaceutical product that was a generic version of the Product, incorporating the patented compound in any dosage form; and “Teva Product” meant any Generic Product marketed by Teva and/or their affiliates.

Clause 2 is headed “Settlement and Covenant not to Sue”. Clauses 2.1 and 2.2 provided as follows:

Clauses 2.3 to 2.5 then provided for AstraZeneca to withdraw and discontinue the various sets of court and arbitration proceedings they had commenced in Portugal within ten days of Teva performing their undertakings in clause 2.1.

Clause 2.6 is a general release in these terms:

Finally, clause 2.7 contains the undertaking by Teva not to sell etc. their generic product before the earlier of certain dates, so far as currently relevant, before 3 July 2017:

At the time that the Settlement Agreement was entered into, the extant intellectual property right of AstraZeneca was SPC 156. However, European patent law recognises the possibility of a six month extension to SPCs by way of a Paediatric Extension. These were designed to address the concern that pharmaceutical products used to treat children were not appropriately authorised. They provide for an extension of six months to the term of an SPC, provided that four conditions are met:

A Paediatric Investigation Plan (“PIP”) is drawn up and approved by the European Medicines Agency;

The clinical trials in the PIP are carried out;

The Paediatric Committee carries out a compliance check to certify that the PIP has been completed; and

The results of the clinical trials are reflected in the product literature.

An application for a Paediatric Extension must be filed no later than two years before the expiry of the SPC on which it is based. The deadline for AstraZeneca was therefore 30 June 2015. At the time of the Settlement Agreement, according to the evidence filed on behalf of AstraZeneca, they had drawn up a PIP which was public, but it was uncertain whether it would be completed in time to meet the 30 June 2015 deadline for a Paediatric Extension. In the event however a Paediatric Extension to the SPC was granted on 30 August 2015 which extended the term of the SPC to 29 December 2017. It is that extension which has given rise to the current dispute.

In reliance on that Paediatric Extension, on 28 April 2017 AstraZeneca commenced arbitration proceedings against Teva in Portugal. From the letter commencing arbitration, it was clear that the claim was to enforce rights arising under SPC 156 as extended by the Paediatric Extension, in particular by preventing Teva from selling their generic product before 29 December 2017. It was against that background that Teva commenced the current proceedings against AstraZeneca and sought relief on an expedited basis and Blair J ordered the summary judgment application to be heard on 30 June 2017.

Following the hearing before Leggatt J, the parties reached an agreement on terms that Teva undertook not to sell etc. their generic product in Portugal before the earlier of 30 December 2017 or the date of determination of this appeal and AstraZeneca undertook to compensate Teva for any damage suffered as a result of giving that undertaking, in the event that the appeal succeeded.

The judge noted the terms of the general release in clause 2.6 but accepted, at [12] and [13] of the judgment, the argument on behalf of AstraZeneca that there was an implicit restriction on the extent of the release, to claims based on the “Patent” as defined in the Agreement. As the judge said at [13]:

The judge went on to accept, at [19], AstraZeneca’s construction of the Settlement Agreement, that “Patent” in the Settlement Agreement meant the basic European Patent and SPC 156 in the form it stood at the date of the Agreement without any Paediatric Extension. Accordingly, he concluded that the subject-matter of the Settlement Agreement was claims and disputes relating to those rights only and not to rights under the Paediatric Extension. His reasoning for reaching that conclusion is in the preceding three paragraphs of the judgment at [16] to [18]:

Teva’s case on the correct construction of the Settlement Agreement is that, by clause 2.6, AstraZeneca irrevocably released Teva from claims “arising from the…offer for sale, sale, marketing…of the Teva Product in Portugal other than in breach of this Agreement.” Since clause 2.7(a) identified 3 July 2017 as the latest specific date from which their undertaking not to sell etc. the Teva Product in Portugal would no longer apply, sale of the Teva Product after that date cannot be a breach of the Agreement. That analysis is not affected by the fact that the term of SPC 156 was extended after the date of the Settlement Agreement by the Paediatric Extension.

In his oral submissions, Mr Maclean QC submitted that the judge’s analysis rejecting this construction and accepting AstraZeneca’s construction that the general release in clause 2.6 did not extend to a claim brought to enforce SPC 156 as extended by the Paediatric Extension, involved three errors of law.

First, he submitted that the judge had lost sight of the principle enunciated by Lord Neuberger at [19] of Arnold v Britton [2015] UKSC 36; [2015] AC 1619:

Mr Maclean QC submitted that the judge had invoked the principle of commercial common sense retrospectively in [18] of the judgment, where he focused on the absurdity of the consequences of Teva’s construction. In other words, his conclusion as to the meaning of the Settlement Agreement at the time it was entered into was driven by the facts as they had turned out. However, at the time the Settlement Agreement was entered into, what might happen was uncertain. Proceedings in Portugal might have continued and AstraZeneca might have lost, with the Patent being invalidated. A third party generic pharmaceutical manufacturer might have succeeded in having the Patent declared invalid, a possibility which was expressly contemplated by clause 2.7. Accordingly, it simply did not follow that, at the time the Settlement Agreement was entered into, AstraZeneca was conferring on Teva a privileged position.

The second error of law which Mr Maclean QC submitted had been committed by the judge was that he had lost sight of the point made by Lord Neuberger at [20] of Arnold v Britton:

Mr Maclean QC submitted that the natural meaning of the definition of “Patent” in the Settlement Agreement encompassed not just SPC 156 as it stood at the date of the Agreement but as it was extended by the Paediatric Extension. Accordingly, the release in clause 2.6 did entitle Teva to sell etc. their generic product after 3 July 2017 and on the correct construction of the Settlement Agreement, AstraZeneca was not entitled to prevent such sales until after 29 December 2017. There was no ambiguity in the meaning of the relevant provisions of the Settlement Agreement and the judge had been wrong to allow considerations of what he had seen as commercial common sense to lead him to reject the natural meaning of those provisions.

The third error of law upon which Mr Maclean QC relied was that he submitted that in the last sentence of [13] of the judgment (which I quoted above), the judge had misapplied the “cautionary principle” derived from BCCI v Ali when he had referred to the type of claims “which the parties may reasonably be taken to have had in contemplation”. Mr Maclean QC referred to [10] of Lord Bingham’s opinion, where he said:

Lord Bingham then set out the various authorities and concluded at [17]:

Mr Maclean QC also referred the Court to passages in the opinion of Lord Nicholls at [27] to [29]:

Mr Maclean QC submitted that this was not a case where a claim or dispute under the Paediatric Extension could be said to be unknown or unforeseen. How far down the line AstraZeneca were in terms of obtaining the Paediatric Extension at the time of the Settlement Agreement does not matter. It was clearly in the contemplation of the parties at the time and the judge had been wrong to conclude that any dispute arising under the Paediatric Extension fell outside the scope of what was settled by the Settlement Agreement.

The primary submission of Mr Ian Mill QC for AstraZeneca was that the judge had been right in his construction of the Settlement Agreement. He submitted that there had been no error of law in the judge’s analysis. Contrary to Mr Maclean QC’s submission, the judge had not been looking at the issue of commercial common sense retrospectively. It was clear from [17] of the judgment that he was looking at the position as at the time that the contract was made.

Equally, he submitted that the judge had not overlooked [20] of Lord Neuberger’s judgment in Arnold v Britton. Here there were two competing reasonable constructions of the relevant provisions of the Settlement Agreement and, as Lord Hodge had confirmed in [11] of his judgment in Wood v Capita Insurance Services Ltd[2017] UKSC 24; [2017] 2 WLR 1095, in those circumstances, the Court was entitled to prefer the construction which was more commercially sensible. In the present case, the judge had not erred in his analysis of which construction was more commercially sensible merely because he had not alluded to the benefits which Teva alleged AstraZeneca had obtained from the Settlement Agreement. This was because the judge had been entitled to take the approach that Teva had no realistic prospect of challenging the validity of the Patent. As Mr Mill QC put it, the settlement of the dispute as it stood in February 2013 had only a nuisance value so far as AstraZeneca were concerned.

He submitted that there had been no error of law by the judge in his application of the BCCI v Ali cautionary principle. The suggestion that the judge had erred in looking at the state of mind and knowledge of both parties at the time of the Settlement Agreement as opposed to just that of AstraZeneca as releasor under the Agreement was misconceived in the light of the opinions in that case and the various authorities cited by their Lordships. In addition to the passages in the opinions of Lord Bingham and Lord Nicholls to which I have already referred, Mr Mill QC referred to what Lord Hoffmann had said at [41]-[42], albeit in a dissenting opinion but in line with the rest of the Appellate Committee on this point:

Mr Mill QC also emphasised the need for circumspection on the part of this Court in relation to the assessment by an experienced Commercial Court judge of the commercial implications of rival constructions of an agreement, with which this Court should be slow to interfere. He submitted that this Court should be particularly circumspect in the context of an appeal against an interlocutory decision on a summary judgment application, relying upon what Beatson LJ said in AmTrust Europe Ltd v Trust Risk Group SpA[2015] EWCA Civ 437; [2015] 1 All ER (Comm) 325 at [37] to [43].

Mr Mill QC’s fall-back position was that, even if this Court was provisionally in favour of Teva on the issue of construction, AstraZeneca had a real prospect of success on their construction at trial when various disputes of fact which remained in issue would be resolved. Accordingly, in those circumstances, this Court should remit the case to the Commercial Court so that it could go forward to trial. He submitted that there were a number of matters of fact, including those referred to in the Respondent’s Notice, which remained in dispute.

He took particular issue with the statement in Teva’s Supplementary Skeleton Argument that part of the factual matrix was the prospect of a defence of invalidity of the Patent succeeding in Portugal and that there had therefore been considerable value to AstraZeneca in having the Settlement Agreement to end the commercial uncertainty of the Patent being under threat in that respect. He relied upon the evidence from his instructing solicitor, Dr Stothers, in a statement dated 2 November 2017 (after the hearing before Leggatt J but which it was agreed in a Consent Order dated 20 November 2017 could be relied upon) to submit that a defence of invalidity could not be raised in private arbitration proceedings in Portugal, which was simply the wrong forum. Whilst Mr Mill QC accepted that such a defence could have been raised by Teva in Court proceedings in Portugal, he submitted on instructions that this defence of invalidity was wholly unmeritorious, although he accepted that there was no evidence before the Court to that effect, other than the assertions by Ms Lacombe of AstraZeneca at [17] to [23] of her witness statement dated 16 June 2017.

Mr Mill QC submitted that, on that basis, the settlement had only a nuisance value for AstraZeneca but that, since there was a dispute of fact as to what, if any, value the settlement had for AstraZeneca, that was a dispute which could only be resolved at trial. If AstraZeneca’s case on the facts was established at trial, it would support their case on construction that Teva’s construction was commercially unreasonable because it would give Teva a windfall of tens of millions of euros during the period July to December 2017, as the judge found at [18], whereas for AstraZeneca the settlement had only a nuisance value.

The other aspect of what Teva asserted was the factual matrix, with which Mr Mill QC particularly took issue, was the statement in the Teva Supplementary Skeleton that neither Teva nor AstraZeneca were unaware of the prospect of a Paediatric Extension. In support of that statement, Teva only relied upon AstraZeneca’s evidence from Ms Lacombe who said at [27]:

Mr Mill QC submitted that if it was true that Teva were aware of the Paediatric Extension, one would have expected them to say so in their evidence, not simply to rely upon Ms Lacombe. Mr Mill QC said that, whilst he would have to accept that as a matter of fact Teva were theoretically aware at the time the Settlement Agreement was entered into that a Paediatric Extension could be applied for, he challenged that they were in fact aware of this Paediatric Extension.

In my judgment, this is not a case where there is any ambiguity as to the meaning of the provisions of the contract or where there are two reasonable interpretations of those provisions. The critical point is that the Paediatric Extension is just that, an extension of SPC 156, prolonging the period of that SPC, not conferring a different kind of right, as the judge himself recognised during the course of argument at the hearing before him. Accordingly, the arbitration proceedings which AstraZeneca commenced in April this year are seeking to enforce rights under SPC 156, albeit during the period of extension. The definition of “Patent” encompasses SPC 156, whether as it originally stood or as extended, so that, contrary to Mr Mill QC’s submissions, the correct construction of the Settlement Agreement is that: (i) AstraZeneca are precluded from bringing a claim in respect of any dispute arising under SPC 156 as extended by the Paediatric Extension; and (ii) the sale etc. by Teva of their generic product in Portugal after 3 July 2017 was not in breach of the Settlement Agreement.

I consider that the judge was wrong to conclude that the definition of “Patent” should be construed as excluding any Paediatric Extension to SPC 156, when the clear and natural meaning of the definition was such as to include any extension to the period of the SPC such as was effected by the Paediatric Extension in the present case. To the extent that the judge was diverted from that clear and natural meaning by considerations of what he saw as commercial common sense, he did fail to have regard to the principle enunciated by Lord Neuberger in [20] of Arnold v Britton. The natural meaning of the provision should not have been subverted by such considerations.

Even if considerations of commercial common sense were relevant, I consider the judge fell into error in his analysis. There is some force in Mr Maclean QC’s submission that the judge was looking at the issue of what was commercially acceptable or otherwise retrospectively when he focused on the consequences in [18] of the judgment. However, even if Mr Mill QC is right in his submission by reference to [17] of the judgment that the judge was looking at the issue as at the time the Settlement Agreement was made, the error he has made is that he has only looked at one side of the overall commercial picture. As Sales LJ pointed out during the course of argument, this is evident from the last two sentences of [16] where the judge appears to have overlooked that Teva had launched their generic product in Portugal prior to the Settlement Agreement and the main commercial point of the Settlement Agreement was to settle the dispute arising out of that by, in effect, preserving AstraZeneca’s exclusivity in Portugal for a period of over four years until July 2017, unless some pharmaceutical manufacturer other than Teva challenged the validity of the Patent or launched its own generic product.

In other words, the judge’s analysis of commercial common sense left out of account the undoubted commercial benefits to AstraZeneca of the Settlement Agreement. As Mr Maclean QC submitted these were: (i) Teva gave up any challenge to the validity of the Patent as defined in any form in any forum in Portugal; (ii) AstraZeneca thereby removed the risk of a finding of invalidity which would have led to them suffering irreparable damage; (iii) AstraZeneca also removed the risk that it would prove impossible to obtain an injunction restraining Teva from selling etc. their generic product in Portugal and (iv) as I have said, AstraZeneca’s exclusivity was preserved unchallenged in Portugal for over four years, unless challenged by someone other than Teva.

Once those commercial benefits of the Settlement Agreement to AstraZeneca are taken into account, it seems to me that the fact that the effect of construing the Settlement Agreement according to its clear and natural meaning is that, after 3 July 2017, Teva were entitled to sell etc. their generic product, even though AstraZeneca had obtained an extension to the period of SPC 156 until 29 December 2017, does not make that construction commercially unacceptable.

In reaching a conclusion as to the correct construction of the Settlement Agreement which is contrary to that of the judge, I have been conscious of the need to accord appropriate respect to the decision of an experienced Commercial Court judge. However, where the issue before this Court is one of the correct construction of a contract, a question to which, as a matter of law, there is only one answer, this Court is entitled to adopt a less restrained approach than would otherwise be appropriate, for example in relation to issues which involve the exercise of discretion by the judge: see Am-Trust Europe at [61] per Beatson LJ and at [72]-[73] per Christopher Clarke LJ.

I was not impressed by Mr Mill QC’s fall-back position that, if this Court was against him on the issue of construction of the Settlement Agreement, we should nonetheless remit the case to the Commercial Court, on the basis that AstraZeneca had a real prospect of success in relation to the issue at trial by virtue of the various matters set out in the Respondent’s Notice and the other matters to which he referred in his submissions. As I have already indicated, the thrust of his submissions on this part of his case was that (i) there was no merit in Teva’s challenge to the validity of the Patent as defined at the time of the Settlement Agreement and there needed to be a trial in England in order to determine that absence of merit; and (ii) there was no evidence that Teva were in fact aware of this Paediatric Extension.

So far as the first point is concerned, I agree with Mr Maclean QC that the suggestion that there has to be a trial before the English Court of the merits of the original dispute in Portugal in circumstances where that dispute was settled by the Settlement Agreement is not only surprising, but does not represent the law. As the judge himself said during the course of argument, where the whole purpose of the Settlement Agreement was to avoid disputes, the suggestion that in order to construe that Agreement it is necessary for the Court to form a view of the rights and wrongs of the underlying dispute, would defeat the whole object of the Settlement Agreement. In any event, whatever the rights and wrongs of the underlying dispute, the Settlement Agreement undoubtedly conferred the commercial benefits on AstraZeneca which I identified in [42] above. It was no answer for Mr Mill QC to rely upon what Ms Lacombe said at [17] to [24] of her witness statement about AstraZeneca’s reasons for entering the Settlement Agreement. This evidence of subjective intention was irrelevant and inadmissible.

In relation to the second point about the knowledge of Teva concerning the Paediatric Extension, given that, through Ms Lacombe’s statement, AstraZeneca not only accepted but asserted that, at the time of the Settlement Agreement, Teva knew about the PIP (which was an essential first step to obtaining a Paediatric Extension), it was not necessary for Teva to adduce evidence to that effect. On the evidence before the Court, both parties knew that it was probable a Paediatric Extension would be obtained in due course. If the disputes between the parties about the validity of the Patent had continued and not been settled, those would have prejudiced SPC 156 as extended by such Paediatric Extension. Assuming in favour of AstraZeneca that for the purposes of assessing whether a particular dispute is unknown or outside the contemplation of the parties, it is necessary to look at what was the knowledge of both parties at the time the Settlement Agreement was made, it cannot be said in this case that a potential dispute arising under SPC 156, as extended by a Paediatric Extension at a later date, was outside the contemplation of the parties.

In any event, irrespective of what was known or anticipated by the parties or by Teva alone at the time of the Settlement Agreement, it is clear from the passages in the opinions of Lord Bingham and Lord Nicholls in BCCI v Ali cited above, that: “the mere fact that a party was not actually aware of the possibility of a claim of the kind in issue is not enough to escape from the terms of a general release”: see per Kitchin LJ in Stretchline Intellectual Properties Ltd v H & M Hennes & Mauritz UK Ltd [2015] EWCA Civ 516; [2016] R.P.C. 13 at [49]. As Kitchin LJ went on to say at [50]:

Applying that principle here, I consider that it is clear that, irrespective of the precise level of knowledge of each of the parties at the time of the Settlement Agreement as to the probability of a Paediatric Extension being successfully obtained, a claim in relation to SPC 156 as extended by the Paediatric Extension related to the particular subject-matter under consideration in the Settlement Agreement, namely disputes relating to the Patent as defined, including the Proceedings, as set out in (vi) of the Preamble. In my judgment, on any view, the general release in clause 2.6 covers such a claim in relation to SPC 156 as extended by the Paediatric Extension as AstraZeneca has sought to advance in Portugal.

In the light of the conclusion I have reached as to the correct construction of the Settlement Agreement, none of the other points in the Respondent’s Notice avails AstraZeneca.

In all the circumstances, for the reasons I have given, I consider that the appeal should be allowed and that Teva are entitled to the declaration they seek.

I agree.

I also agree.