Novartis AG v Dexcel-Pharma Ltd

Novartis AG (“Novartis”) is the proprietor of UK Patent No 2 222 770 (“the Patent”). Novartis seeks an interim injunction to restrain Dexcel-Pharma Limited (“Dexcel”) from marketing a cyclosporin-containing formulation which it alleges infringes the patent. Novartis’ product made in accordance with the Patent is recognised as a significant breakthrough in the treatment of transplant patients. The case is unusual because the Patent has been the subject of previous judicial consideration. Dexcel says that, on the construction previously arrived at, there is no serious question to be tried on the issue of infringement. Dexcel raises no novelty/obviousness challenge to validity, but Dexcel has indicated it may seek to raise an insufficiency squeeze if so advised.

The Patent is entitled “Pharmaceutical compositions containing cyclosporins”. The cyclosporins are a class of drugs possessing immunosuppressive activity. Cyclosporins are important in the treatment of transplant patients, to prevent the rejection of transplanted tissue. A problem with cyclosporins is their poor solubility in aqueous media leading in turn to poor bio-availability.

The Patent proposes a novel cyclosporin formulation in the form of “a micro-emulsion pre-concentrate”. Claim 1 claims

a hydrophilic phase,

The reference to an “oil-in-water micro-emulsion pre-concentrate” refers to the fact that what is claimed in claim 1 is a concentrate which, when added to water, either in a glass so that it can be drunk, or in the human body from a capsule, forms a micro-emulsion. An important feature of a micro-emulsion mentioned in the Patent is that its particles are less than 200 nanometres in diameter.

In the previous litigation to which I have already referred (Novartis v Ivax [2006] EWHC 2506 (PAT); [2007] EWCA Civ 971), Pumfrey J held the Patent to be valid, but not infringed by the formulation intended to be marketed by the generics company Ivax. The Court of Appeal upheld the finding of non-infringement. Two issues of construction and non-infringement arose. The first was that Ivax claimed that its cyclosporin was largely contained in particles which were larger than 200 nm. Jacob LJ, who gave the leading judgment, said

So later he said:

On that basis the IVAX formulation was outside the claim. That was enough to dispose of the appeal. Nobody questions that construction in the present case. But Pumfrey J had held that the Ivax composition did not infringe on a further ground. The second point was whether the Ivax product had a lipophilic phase. Novartis alleged that the lipophilic phase was constituted by polyglycerol-3-oleate (Pg3o). The hydrophilic phase was ethanol. Pg3o is soluble in ethanol, at least in the relative amounts present in the composition.

A particular problem with the allegation of infringement in the Ivax case was that the ingredient in question was also a surfactant, which is another feature of the claim. There were other surfactants present apart from Pg3o. Pumfrey J held at [36]:

The learned Judge went on to say:

Support for Pumfrey J’s construction is to be found at page 15 of the specification where it is said that

Ivax’s formulation thus fell outside the claim for this second reason, because there was no “separately identifiable” lipophilic phase. And it seems this was so in pre-concentrate and aqueous medium alike.

The Court of Appeal also briefly dealt with this second ground. Jacob LJ dealt with the point in this way:

Thus far Jacob LJ was saying that a component like Pg3o could not be the lipophilic phase because it would not do the job expected of that phase in the patent. He went on to deal with the passage in the patent cited above about the miscibility of the lipophilic phase and the hydrophilic phase, rejecting a submission that the passage was not dealing comprehensively with the purpose of the lipophilic phase:

The Court of Appeal was accordingly considering whether the “lipophilic phase” claim element was apt to cover a material which was eminently unsuitable to fulfil the purpose of that element, and which was in fact suited to fulfil the purpose of a different and contrasting claim element, the surfactant. It was in this context that the Court held there must a distinct lipophilic phase. It is important to note that neither Pumfrey J nor the Court of Appeal were, so far as I can see, directing their minds specifically to the question of when the two phases must appear: it does not appear that that was how the argument was presented. To the extent that there is material in the judgments touching this question, it seems to me that they were looking forward to the function of the required phases in the aqueous micro-emulsion.

It is sometimes said that a finding about the correct construction of a given patent specification is a question of law, binding on a subsequent court according to the rules of precedent. Terrell on Patents (16^th edition) at paragraph 16.43 puts it in this way:

That judgment was given in February 1889. Three months later in May 1889, in Automatic Weighing Machine v The Combined Weighing Machine (1889) 6 RPC 367 at 370, a Court of Appeal which included Cotton LJ refused to allow a patentee to reargue the construction of a patent it had previously construed, holding the alleged infringement not to be materially different from that considered in the previous case.

In SmithKline Beecham v Apotex [2004] FSR 26; [2003] EWHC 2939 Pumfrey J was faced with a patent which had previously been construed by the Court of Appeal. He held at [12] that the findings of the Court of Appeal had no bearing on his case, “save on questions of construction”. He said:

So, in that case, where the claim called for “conventional drying conditions”, Pumfrey J felt able to revisit the conclusion reached by the Court of Appeal in the earlier case as to what those conditions (in terms of temperature etc at the relevant date) would be understood by the skilled person to be: see [30] to [34].

When that case reached the Court of Appeal, there was no detailed discussion of this aspect of Pumfrey J’s judgment. But at [4] Jacob LJ said this:

At [115] Jacob LJ said that the fact that Pumfrey J reached a different conclusion about “conventional drying” raised “a very substantial doubt as to whether the claims had any real meaning at all”. But he did not doubt the proposition that evidence in a subsequent case might affect the meaning.

The problem with treating questions of construction as pure questions of law is that they frequently are not. Although direct evidence as to the meaning of non-technical terms is inadmissible, the court does not reach its conclusion as to the meaning of the claim in a factual vacuum. Evidence as to the common general knowledge can frequently have an important bearing, as part of the factual matrix against which construction is decided, as can factual evidence about the consequences of the teaching of passages in the specification. The Ivax case is a good example of both points. The fact that micro-emulsions were held not to form part of the common general knowledge (a fact heavy finding) can be seen to have influenced Pumfrey J’s conclusion on construction. Equally, the absence of evidence as to the consequences of certain examples in the specification was material to the Court of Appeal’s rejection of arguments on construction e.g. at [27]. Also, as SKB v Apotex shows, the meaning of a claim feature may be dependent on external evidence about what it conveys which may change from case to case.

Thus while a subsequent court is bound on questions of law, care is necessary to identify the precise point of law to be derived from the previous court’s decision. For example it must be that, faced with evidence about the common general knowledge and the disclosure of the document to the skilled person which is indistinguishable in any relevant way from the evidence in the Ivax case, the same conclusion about construction must follow as a matter of law in a subsequent case. But if the evidence on which the legal conclusion is founded is different in a relevant way, one may ask legitimately, why should subsequent litigants be bound by the evidence adduced in a previous action? The point of law to be decided once the evidence is different is itself different. I think this comes within what Jacob LJ had in mind in SKB v Apotex when he said that the subsequent court is not bound by the previous court’s decision “so far as it turned on evidence”.

For present purposes – this is an interim application after all - all I need say is that I regard it as seriously arguable that a subsequent court is not bound by a previous court’s conclusion as to the correct construction of a patent specification if that conclusion turns in a material respect on evidence which was either lacking in the previous case or which is materially different in the present one.

There is no real dispute about the principles which I should apply when deciding whether to grant an interim injunction. In deciding whether there is a serious issue to be tried, I should not embark on anything approaching a mini- trial on the basis of expert statements without cross-examination. If I am satisfied that there is a serious issue to be tried, I should proceed to determine where the balance of convenience, sometimes called the balance of injustice, lies. Where the balance of injustice appears evenly balanced, it may be appropriate to take steps to preserve the status quo. Equally, as Lord Diplock said in American Cyanamid v Ethicon [1975] AC 396 at 409

The need only to show a serious issue to be tried should not be taken to indicate that the court will proceed on hypotheses of fact. The burden is on the party seeking the relief to adduce evidence of a sufficiently precise nature to satisfy the court that it has a real prospect of succeeding in obtaining a permanent injunction at trial: see per Slade J in Re Lord Cable [1976] 3 All ER 417 at 430. In the context of a patent case the evidence must be sufficiently precise to enable the court to come to a view that there is a serious prospect that the patentee will be able to show that each of the features of the claim is present in the alleged infringement.

The product which Dexcel wishes to launch is called DEXIMUNE, and is described in Dexcel’s PCT Application and regulatory particulars. It is common ground that it is a pharmaceutical composition comprising a cyclosporin as its active ingredient. I consider each of the claim features and the position on the evidence in turn.

The hydrophilic phase: DEXIMUNE contains ethyl lactate. Novartis alleges that this constitutes the hydrophilic phase. Ethyl lactate is described as a hydrophilic solvent in Dexcel’s PCT application. Dr Mullertz, Dexcel’s expert, does not agree that this is a complete description of ethyl lactate, but does not really challenge that it is correct to describe it as hydrophilic. I will have to return to whether it is a phase.

The lipophilic phase: DEXIMUNE contains tricaprin which is a medium length chain fatty acid triglyceride. It is a member of the class of preferred candidates for the lipophilic phase in the Patent. This is a significant difference from the IVAX case, where the lipohilic phase was said to be constituted by an amphiphilic surfactant. There is no suggestion that the lipophilic phase in this case will be incapable of doing its job as carrier of the cyclosporin and forming the dispersed phase in the aqueous dispersion. Professor Lawrence, Novartis’ expert, has expressed the view that a significant amount of the cyclosporin will be carried in the tricaprin. Whether this view turns out to be correct is a matter which can only be determined at a trial.

The surfactant: DEXIMUNE has a surfactant. It is one described in the Patent as particularly suitable: polyoxyl 40 hydrogenated castor oil.

The miscibility point: It appears to be common ground, at least for the purposes of this application, that ethyl lactate and tricaprin are miscible, at least in isolation. This is, accordingly, the strongest of Dexcel’s non-infringement arguments on this application, because if miscibility is a ground of non-infringement, the facts are uncontested. Dexcel draws the comparison with the Ivax case. If the two components are miscible, there can be no separately identifiable phases in the pre-concentrate. That is what Pumfrey J and the Court of Appeal were saying, Dexcel submit, was essential in the pre-concentrate.

Novartis submit that there is no technical reason why the skilled reader would regard it as important to have two separate phases present in the pre-concentrate: what is important is what happens when the pre-concentrate is dispersed in aqueous solution, either in a glass or in the stomach. That is the purpose for which the components exist. They rely on Professor Lawrence again (in her reply report at [13]):

I did not understand the evidence of Dr Mullertz for Dexcel specifically to take issue with this evidence, or to suggest there would be any technical reason why one would seek to have two separate phases present in the pre-concentrate as opposed to the ultimate micro-emulsion.

Dexcel submit that this is an attempt to revisit or undermine the construction placed upon the claim by Pumfrey J and the Court of Appeal. I think there is an answer or at least a potential answer to that submission. Professor Lawrence’s evidence is directed to the issue of when not whether two separate phases are necessary. It is clearly arguable that that issue is not one which was formally concluded by the Court of Appeal’s judgment. My reading of the Court of Appeal’s judgment is that they were focusing on Novartis’ more radical proposition, that there was no need for those distinct phases ever to exist at all. Even if that is wrong, it is arguable that, if Professor Lawrence’s evidence is accepted to represent the technical position at trial, it will be open to the trial judge to reach a different conclusion for the reasons I have endeavoured to explain above.

It also seems to me that there is a stark difference between the point in the Ivax case and the point here. In Ivax, the surfactant was incapable of fulfilling the purpose of that component and would be considered to fall outside the claim. But it is a quite different thing to ask whether a component which is lipophilic and is capable of forming the oil part of the dispersion is within the claim. Here considerations of purpose would seem to favour a finding of infringement, not militate against it. Jacob LJ’s characterisation of the fundamental purpose - avoid the problems of the prior art emulsion formulation by carrying the cyclosporin in the micro-emulsion size particles – would seem to be met by Dexcel’s proposed formulation on Novartis’ case, in stark contrast to that in Ivax.

Moreover Novartis seeks to rely on further evidence that was not available to either Pumfrey J or the Court of Appeal, namely the fact that a large proportion of the examples in the Patent employ a combination of lipophilic and hydrophilic components which are miscible in isolation. That evidence, although exhibited by Novartis, in fact comes from Dexcel who deployed it in support of an insufficiency argument elsewhere. Novartis rely on it to show that if immiscibility per se is taken to be a requirement of the claim, a very large proportion of the favoured embodiments of the invention will be held to be outside the claim.

Mr Thorley’s answer to the latter point is to say that it is not established that the examples in Novartis’ patent in fact produce micro-emulsions. This may conceivably turn out to be correct; but I would regard the patent itself as providing adequate evidence for present purposes that micro-emulsions are formed in the examples as the patent promises.

Mr Thorley also argued that even if it turned out that the examples of miscible combinations in the Patent do as they say and form micro-emulsions, the skilled person would still understand the claim as being restricted to separate phases in the pre-concentrate. He relies strongly on Pumfrey J’s conclusion that the art of forming micro-emulsions was not common general knowledge. The skilled person would have no choice, he says, but to follow literally the recipe he is given in the Patent and to ensure his ingredients are ones that are immiscible in isolation.

As he developed it, this is a formidable submission. I confess that at one point I was impressed by the argument that the claim to pre-concentrate requires phases to be present there as well as in the ultimate dispersion. That may, however, be taken to be too literal a view, and not to give sufficient weight to the patentee’s purpose. Moreover, I think it is arguable that the Patent itself teaches the reader enough about micro-emulsions to know that what will be important will be miscibility in aqueous solution. That is, after all what the patent would be seen to be driving at. Moreover, I would be reluctant to hold that Novartis are precluded by the findings in the Ivax case from contending that the specific examples in the Patent are covered by the claims – that strikes me as a potentially unjust, as well as an unusual result.

The final question under this heading is whether Novartis has adduced sufficient evidence to show that separate phases are formed at any stage. Novartis has performed an experiment to show that there is phase separation. Mr Thorley has severe criticisms to make of them; but Professor Lawrence’s evidence was that the experiments do suggest that in an aqueous dispersion of ethyl lactate and tricaprin are largely immiscible and form two distinct phases.

Accordingly I would hold that Novartis’ case on the miscibility point is seriously arguable.

The micro-emulsion: It is still necessary for Novartis to show with sufficient precision for an interim application that a micro-emulsion is in fact formed when the Dexcel preparation is placed in an aqueous environment.

Novartis start from the proposition that Dexcel itself so describes it in its PCT and regulatory documents. For example its Public Assessment Report - Scientific discussion” says

Furthermore, and unlike Ivax, Novartis’ evidence is that some 97% of the particles are of micro-emulsion size. The example in Dexcel’s PCT application shows that the product had an average particle size of 25-50 nm. So it is a reasonable inference that a significant amount of cyclosporin will be present in the smaller particles.

Mr Thorley also relies on Pumfrey J’s conclusion in the IVAX case at [25] that a micro-emulsion has a number of characteristics including that they are thermodynamically stable, form spontaneously and transparent or translucent. At [28], Pumfrey J held that contrary to the teaching of the specification micro-emulsions must possess all of these characteristics, not just one or more of them. I am not certain that this can be characterised as a finding about construction at all. For example the Court of Appeal was not content to take translucence as a determinative requirement of a composition of the invention: see [20]. Moreover they did not think that either party would get much out of the relevant passage in the specification. Even if Mr Thorley is right, and there is a series of further sub-requirements of a micro-emulsion, I am not persuaded that Novartis has failed to demonstrate an arguable case in relation to any of them. Thus, in addition to particle sizing, the parties have also done opacity tests. But this can only be a rough way of showing whether a micro-emulsion is present because the opacity might be due to factors other than large particles: or it might be due to the very small proportion of larger particles. In any case, the evidence would require expert interpretation (Novartis says it is translucent and Dexcel says opaque) and there is a dispute about how far one should dilute. Mr Thorley had other criticisms of the evidence on spontaneous formation of the micro-emulsion and stability: but all this is plainly in issue and I could not fairly resolve it now.

In my view there is a serious question to be tried as to whether DEXIMUNE infringes claim 1 of the Patent.

Firstly, the timescale to trial. I have made a directions order which would enable a trial to take place in late 2008 or early 2009. I was invited by both parties to assume that this is the approximate timescale over which an interim injunction, if granted, would run.

Secondly, the chronology to date. Novartis learned that Dexcel had been granted a conditional marketing authorisation in the UK in June 2007. Novartis wrote to Dexcel at that time, warning it that its product might infringe and that the Patent had been held valid by Pumfrey J subject to appeal. It was noted for the record that, despite the marketing authorisation, Dexcel had made no attempt to clear the position with Novartis so far as UK patent rights were concerned.

Dexcel’s response was not to make any attempt to clear the way so far as patent rights were concerned. Instead on 12^th June 2007 it gave an undertaking until 13^th July not to launch, and on 16^th July it stated:

There was no further correspondence between 16^th July 2007 and 14^th March 2008 when Dexcel’s solicitors served the notice pursuant to the previous arrangement. The letter drew attention to the judgments in the Ivax case (the Court of Appeal had now given judgment as well), and to decisions in Denmark in proceedings against Actavis A/S both at first instance and on appeal where a claim against an identical product to DEXIMUNE had failed. It was this letter which prompted the present application.

In fact, through no fault of Dexcel, the marketing authorisation is still not finalised. Dexcel expects it to be finalised in the next 60 days, leading to a launch some time in August this year.

There are a number of points to make about this chronology. Firstly, Dexcel can have been in no doubt from early 2007 that Novartis would sue if it launched DEXIMUNE in the UK. So much was clear from the steps Novartis had taken against Actavis and other competitors. Secondly, if it entertained any such doubt, then it should have taken the matter up with Novartis. It did not do so. What it should have done is put Novartis on the spot, by stating its view that marketing would not infringe. Alternatively it could have sought a statutory declaration of non-infringement. Far from doing that, it told Novartis it had no immediate intention to launch and gave Novartis an undertaking which effectively precluded Novartis from taking action against it. What the undertaking clearly did not preclude was an action by Dexcel to seek a declaration of non-infringement. Accordingly, if one asks a question about whose responsibility it is that the legal position is not clear before Dexcel’s planned launch date, the answer is very clearly that it is Dexcel’s. Thirdly, the period we are considering during which an interim injunction would bite is from a launch in August 2008 to a trial in early 2009.

Next, I should consider the position on loss to the respective parties on the assumption that I take the “wrong” decision now. I consider the position of Novartis first of all.

Novartis currently enjoys a monopoly position in micro-emulsion formulations of cyclosporin. Moreover, Ivax has not yet launched its non-infringing product, despite the favourable Court of Appeal decision last year. There is no evidence as to when it might, although it is apparently on the market elsewhere. It appears that Ivax may not yet have succeeded in clearing the regulatory path. If it enters the market between now and trial the computation of damages will be rendered significantly more complex.

NEORAL, which is Novartis’ product and the embodiment of the patented invention, represents 7% of Novartis’ turnover in the UK. Novartis’s evidence is that generic entry would force a downward price spiral or significant loss of market share for NEORAL. Novartis will be placed in the familiar dilemma of either having to reduce its price or lose market share. Ms Mary-Lynn Manning, who is Novartis’ Director of Infectious Diseases, Transplantation and Immunology and its deponent on this issue, says that she does not expect as dramatic a fall in the present case as has been experienced with certain other drugs, because physicians show a reluctance to change a patient from one drug to another for such a critical condition as transplant surgery rehabilitation. But the opposite side of that coin is that it may involve a loss of goodwill for Novartis to attempt to recapture patients started on DEXIMUNE when it succeeds (on this hypothesis) at trial.

60% of NEORAL sales are sold by branded prescriptions: so the pharmacist can only dispense NEORAL even if there is a generic drug available. Novartis therefore enjoys some protection for that portion of the market even if there is competition. If DEXIMUME is launched it will be a branded generic according to the evidence.

Ms Manning also considers that the recovery of Novartis’ market position following a successful decision at trial would give rise to difficulties. In particular, the loss of goodwill that would follow from enforcing an increase in price would make it commercially impossible to recover the former price position. As a result Novartis would lose a significant portion of the value of the remaining life of the Patent, which has only another 18 months to run. I think there is force in these considerations.

Novartis also fears that refusal of the injunction would unleash further unlawful competition. There is some evidence of other companies with products waiting to come onto the market, but which have presently undertaken not to do so without notice to Novartis. This could intensify the price war. Again, there is some force in this, although the potential for too much of this to happen before early 2009 should not be overstated.

Next I consider the position of Dexcel. DEXIMUNE is clearly a strategically important product for Dexcel, as its deponent Mr Kappler, its Managing Director, explains.

Dexcel understandably wish to be the first generic on the market. It is well recognised now that the first generic entrant gains more of the market than subsequent entrants, and holds on to it. If Dexcel is restrained, it is possible that Ivax will beat it to the market. Moreover if it is relegated to third place by Ivax, then the effect on its business will continue for many years at a level below that at which it would have been if it had been allowed to launch now. The effect of that will be hard to quantify as well.

Both sides recognise that the other will suffer a measure of unquantifiable loss if I make the “wrong order”. Balancing the risk of injustice is made very difficult by the uncertainty as to whether Ivax will launch. On the whole I think there is more to put in the scales on Novartis’ side than on Dexcel’s. Moreover, I am impressed by the fact that Dexcel had it within its power to clear its product in time for its launch and neglected to do so. The fact that it claims the product to be a strategic and important one makes its strategy all the more surprising.

Weighing all these factors, I consider the balance of convenience to favour Novartis. It seems to me that even if I were not persuaded that the balance of justice were in favour of Novartis, and that it was more or less evenly balanced, I would nevertheless have thought this an appropriate case in which to preserve the status quo. The status quo is that there is no infringing product on the market. I do not feel able, on the necessarily incomplete evidence before me, to come to a sufficiently clear view on the merits as disclosed in the witness statements to allow that to influence the result.

I should add that I have treated both sides as equally good for the damages they might have to pay. There was some dispute as to whether Dexcel-Pharma Limited was good for any damages that might be awarded. Dexcel offered to make a bank deposit in a sum of £2million or such greater sum as the court thought appropriate. In the light of the conclusion I have been able to reach it is not necessary to take any of that into account.

I have come to the conclusion that this is a case where, in the probably short period between now and trial, it is appropriate to grant the interim injunction sought by Novartis.