Novartis AG v Hospira UK Ltd

This is an appeal from the judgment of Birss J dated 14 May 2013 by which he refused Novartis AG an interim injunction in a patent action pending an appeal to this court. The appeal raises a question about the grant of interim injunctive relief pending appeal when the patent in suit has been held invalid. We heard the appeal on an urgent basis on 17 May 2013 in view of Hospira UK Limited’s intention imminently to launch a generic drug which is alleged to infringe the patent.

The appeal arises in the following circumstances. Novartis AG is the proprietor of two European patents (UK) Nos. 1296689 and 1591122 which cover the use of zoledronic acid for the treatment of osteoporosis. These patents both have a number of years to run. I will call them “the use patents”. The use patents cover Novartis’ drug Aclasta. Worldwide sales of Aclasta were some US$590 million in 2012, of which about £5.7 million represented UK sales. The market is said by Novartis to be an expanding one.

Birss J described the invention of the use patents in this way:

“The point of the invention is that zoledronic acid can be administered as a single intravenous injection half-yearly or even annually. Other older bisphosphonate compounds in this class are administered orally, which would seem to be more convenient than by injection. However, the problem with oral administration is gastric intolerance. Thus, when a patient takes the tablets, they have to remain standing for half an hour after taking the tablets to seek to mitigate the side effects. The tablets also have to be taken relatively frequently.

Although it was known that intravenous administration would avoid the gastric side effects, the problem was that it was thought that the bisphosphonate injections would have to be given frequently, which is also undesirable. The invention in this case was based on the discovery that zoledronic acid, a known bisphosphonate, could be effective when administered intravenously half-yearly or yearly.”

Novartis is also the proprietor of an earlier patent and corresponding SPC covering zoledronic acid itself. The SPC expired on 15^th May 2013, that is to say in the course of last week. I will refer to it as “the SPC”. Whilst the SPC was in force no one could sell zoledronic acid for any purpose, let alone for intravenous use for the treatment of osteoporosis.

On 20 December 2011 Hospira, a well known and well respected generic medicine manufacturer and supplier, commenced revocation proceedings against the use patents (“the revocation proceedings”). The revocation proceedings were commenced, so far as Hospira were concerned, because they had it in mind to market a generic zoledronic acid product. The revocation proceedings were not expedited in any way, and a trial date, some 14 months from the issue of proceedings, was fixed for February 2013. That is a normal or slightly attenuated period for revocation proceedings to come to trial in the Patents Court.

In November 2012, before the revocation proceedings had come to trial, but after the trial date had been fixed for February 2013, Hospira obtained a marketing authorisation for zoledronic acid for amongst other things osteoporosis. The marketing authorisation also covered the use of zoledronic acid for the treatment of Paget’s disease, a fact which has some relevance to one of the arguments about the balance of convenience.

The fact of Hospira’s marketing authorisation came to Novartis’ attention at the end of November 2012 or early December 2012. Until that date Novartis were not in a position to take any steps to obtain an interim injunction. When the marketing authorisation did come to their attention, Bristows, the solicitors acting for Novartis in the revocation proceedings, wrote promptly to Hospira on 4 December 2012. The letter asked for:

Bristows also said in that letter that if the acknowledgment on the issue of infringement was not given, then the issue of infringement should be heard at the trial in February.

Taylor Wessing, the solicitors acting for Hospira, replied on 14 December 2012. They gave the undertaking which had been requested in relation to the SPC, thus securing Novartis’ position until 15 May 2013. Taylor Wessing resisted the introduction of infringement issues into the February trial. Those issues should, according to them, have been raised earlier.

Bristows wrote again on 18 December 2012. This is a letter to which the judge attached considerable significance. The whole of the letter needs to be read to understand the context. In the first paragraph Bristows recorded Hospira’s undertaking based on the SPC not to sell any zoledronic acid product before 15 May 2013. In the next paragraph they took issue in strong terms with the suggestion that they should have raised infringement earlier, given that they were unaware of any marketing authorisation, amongst other things. The third paragraph deals with the position after 15 May 2013 and was in these terms:

“So that our client knows where it stands, we request that you confirm that, if any of the claims of [the use patents] asserted as having independent validity is held valid by the Court following the trial in February, your client will undertake not to launch its Zoledronic Acid Hospira 5 mg/100ml solution for infusion medicine in the UK after 15 May 2013.”

Such an undertaking was necessary because a finding of validity of the use patents would not of itself prevent Hospira from launching the product covered by its marketing authorisation in the absence of an acknowledgment from Hospira that their product would infringe. The letter was not addressing itself to the question of what would happen if the patents were held invalid.

Hospira continued to resist the introduction of any issues of infringement into the revocation proceedings. On 10 January 2013 Taylor Wessing wrote that

“Hospira wishes to review which products and indications it may market in the future in the UK once it knows the outcome of these proceedings”.

In the end the issue of infringement was not added to the revocation proceedings. The parties agreed that the correct time to address the issue was when judgment was handed down in the revocation proceedings.

The trial of the revocation proceedings took place between 20 and 26 February 2013 before Arnold J. Arnold J’s judgment was handed down on 15^th March. He held both of the use patents invalid. This was on the ground that although the allegations of obviousness failed, the patents were not entitled to their claimed priority date, and hence were rendered invalid by an intervening prior publication. He also upheld an allegation of insufficiency in relation to some but not all of the claims.

On 18 March 2013, following the handing down of judgment, Bristows wrote to Taylor Wessing saying that they were concerned that in the light of the judgment Hospira would be introducing their zoledronic acid solution after 15^th May 2013, that is to say after the date of expiry of the SPC, which of course at that date still remained an obstacle to the introduction of any zoledronic acid product. They asked Taylor Wessing to let them know within 7 days whether it was Hospira’s settled intention to launch on or shortly after 15 May 2013.

Despite reminders, Hospira did not answer this question. Instead, on 4 April 2013, Taylor Wessing wrote asking whether Novartis intended to seek an injunction pending appeal under either or both of the use patents. They pointed out that they did not accept that Novartis would be entitled to an injunction given the finding of invalidity. Bristows replied the next day saying that they believed Arnold J’s approach to priority and insufficiency had been wrong. They said it was indeed their client’s intention to seek an interim injunction against any company which threatened to launch an infringing formulation. They offered a standard form of cross undertaking. They also enquired whether Hospira would support an application by Novartis for the appeal to be expedited. Taylor Wessing replied on 10 April saying that the enforceability of the patents and the timing of any appeal depended on the application for permission to appeal which had not yet been heard.

At a further hearing on 12 April Arnold J gave permission to appeal. In giving permission he said that the grounds of appeal just about persuaded him that an appeal would have a real, as opposed to a fanciful, prospect of success.

On 24 April Taylor Wessing wrote to Bristows saying that it was Hospira’s intention to launch in the UK “following expiry of the SPC on 15 May 2013.” On the same day, in a further letter, they wrote that they would not support any application to expedite the appeal and were considering opposing it.

Accordingly on 26 April Novartis commenced infringement proceedings, by a separate action (“the infringement proceedings”). The issue of the infringement proceedings was accompanied by an application for an interim injunction. A timetable was proposed for evidence which led to an effective, opposed hearing before Birss J on 13^th May 2013. Having heard the argument in the afternoon, Birss J delivered his judgment orally the following day. He declined to grant the injunction sought. Hospira gave an assurance that they would not take any steps to launch the product before 4pm on 17 May.

I granted permission to appeal from Birss J’s judgment on 16 May. This court was convened at short notice to hear the appeal on 17 May, given its urgency.

The judge considered the principles to be applied to an application for injunctive relief pending appeal at [25] to [27] of his judgment. A summary of his conclusions is as follows:

The judge then turned to the question of whether there was a good arguable case. He referred to Arnold J’s remarks on granting permission to appeal. His view was that the appeal was “plainly arguable”, by which I believe he meant that the appeal satisfied the “good arguable case” criterion which he had applied. He added that he could not conclude who would win, and that he was perhaps differing (in the sense of being more optimistic about Novartis’ prospects) from the view expressed by Arnold J on the grant of permission to appeal.

Birss J then turned to the balance of unquantifiable harm. The arguments were the familiar ones in the pharmaceutical patent field. On Novartis’ side it was maintained that Novartis would suffer harm from the effect of Hospira and other generic companies undercutting its monopoly price. It would have to reduce its prices or lose market share. It would be difficult to raise its prices again if successful on appeal. If it did so, it would face damage to its reputation. On Hospira’s side, Hospira would or might lose the advantage of being the first to market, the so-called “first mover advantage”. The first mover advantage is the advantage that the first generic to market enjoys when it enters the market at a price near the monopoly price in the absence of other generic competition. In such circumstances it can reap far greater rewards than second and subsequent companies who would be likely to cause an uncontrolled downward price spiral.

The judge concluded:

At [36] the judge said:

“Looking at uncompensatable harm overall, I am quite satisfied that taking either course will lead to a risk of significant unquantifiable loss. If this case was at the outset of proceedings months or even a year before trial, then as long as all I could say about the merits was what I can now say about the merits of the appeal, I would probably grant the interlocutory injunction. On its facts this is a case like many others in the pharmaceutical field. The magnitude and significance of the unquantifiable harm is such that the best way of holding the ring pending trial might well be to grant an interim injunction. That is where the idea of clearing the way comes from. The generic knows more about its plans than the innovator, so the generic ought to arrange things to have the merits tried before it launches.”

So the judge was saying that, on his assessment of the factors explained in American Cyanamid v Ethicon [1975] AC 396, the case for an interim injunction pending trial would be likely to have been made out. The judge then turned to the balancing exercise before him:

“However, now that the merits have been decided I believe the overall balance is different. If they win the appeal Novartis’ lawful monopoly in zoledronic acid will be restored after a lapse of about 6 to 8 months; I should say 6 to 8 months is what I understand is the likely listing of this appeal. They will have suffered loss which will be a large sum in financial terms, but the damages they recover can all be paid by Hospira and any other generics who launch in the period. Quantification is not at all easy, but an assessment can be made.”

Thus far the judge had not explained what was different in the scenario which he was now considering from that pre-trial. However, at [38] he said this:

“These proceedings have been conducted on the footing that the trial on the judgment of first instance would be concluded before the SPC expired. The point was obviously to clear rights which might subsist after the SPC expired out of the way. If Novartis had wanted to say that the case should be scheduled to allow an appeal to the Court of Appeal to take place before 15th May, that could have been sought a long time ago; it was not. The letter of 18th December was clearly written on the basis that the trial was going to determine the position vis-a-vis launching after 15th May. Novartis argued that the letter was written on an assumption that the patent was valid. I do not accept that that takes away the significance of the letter. In my judgment, it shows that these proceedings have been arranged, as I say, so that the trial will be in a position to have been decided by the time of the launch date on 15th May.”

The judge concluded at [39] that if the position were that the appeal would be rendered nugatory, then the prospects of success which he had found (“plainly arguable”) would have justified an interim injunction. More significantly for present purposes he said that if the appeal had strong prospects of success that might tip the balance in favour of an injunction. As he was in neither position, but instead faced with an arguable appeal and a risk of inflicting significant uncompensatable harm on either side, the risk should be borne by Novartis, the appellant.

Novartis appeal against that decision on the following principal grounds:

There is no dispute that the court has a jurisdiction to grant interim protection to a party who has been unsuccessful at trial pending an appeal. Such interim protection can arise in two ways. An unsuccessful defendant may be granted a stay of execution pending the appeal. Such stays are not granted automatically, but may be granted where solid grounds are shown and subject in appropriate cases to conditions. Similarly an unsuccessful claimant may be granted interim protection if he is seeking to restrain some irreparable harm pending appeal, notwithstanding that he has been unsuccessful in asserting his right at trial.

In Minnesota Mining v Johnson and Johnson [1976] RPC 671 the Court of Appeal set out the principles which are to apply where the claim succeeds and the grant or stay of injunctive relief is being considered pending appeal. At page 676 lines 7-13, Buckley LJ said this:

“It is not in dispute that where a plaintiff has at first instance established a right to a perpetual injunction, the court has a discretion to stay the operation of that injunction pending an appeal by the defendant against the judgment. On what principles ought such a discretion to be exercised? The object, where it can be fairly achieved, must surely be so to arrange matters that, when the appeal comes to be heard, the appellate court may be able to do justice between the parties.”

Buckley LJ went on to discuss the various measures open to the court in the case of a successful plaintiff:

“Where an injunction is an appropriate form of remedy for a successful plaintiff, the plaintiff, if he succeeds at first instance in establishing his right to relief, is entitled to that remedy upon the basis of the trial judge’s findings of fact and his application of the law. This is, however, subject to the defendant’s right of appeal. If the defendant in good faith proposes to appeal, challenging either the trial judge’s findings or his law, and has a genuine chance of success on his appeal, the plaintiff’s entitlement to his remedy cannot be regarded as certain until the appeal has been disposed of. In some cases the putting of an injunction into effect pending appeal may very severely damage the defendant in such a way that he will have no remedy against the plaintiff if he, the defendant, succeeds on his appeal. On the other hand, the postponement of putting an injunction into effect pending appeal may severely damage the plaintiff. In such a case a plaintiff may be able to recover some remedy against the defendant in the appellate court in respect of his damage in the event of the appeal failing, but the amount of this damage may be difficult to assess and the remedy available to the appellate court may not amount to a complete indemnity. It may be possible to do justice by staying the injunction pending the appeal, the plaintiff’s position being suitably safeguarded. On the other hand it may, in some circumstances, be fair to allow the injunction to operate on conditions that the plaintiff gives an undertaking in damages or otherwise protects the defendant’s rights, should he succeed in his appeal. In some cases it may be impossible to devise any method of ensuring perfect justice in any event, but the court may nevertheless be able to devise an interlocutory remedy pending the decision of the appeal which will achieve the highest available measure of fairness. The appropriate course must depend on the particular facts of each case.”

Servier v Apotex (unreported, 6 July 2007) was a case where the patentee’s claim failed, but an interim injunction had been in force pending the trial. No approved copy of the judgment of Pumfrey J exists, but we have been provided with a transcript of what he said. In fact the claim in that case had failed to such a degree that Pumfrey J thought there was no real prospect of success on appeal (a view with which the Court of Appeal subsequently agreed). Pumfrey J nevertheless granted permission to appeal, basing himself on some observations of Jacob LJ about practice in such cases in Pozzoli v BDMO SA [2007] EWCA Civ 588 at [10]. At the passage bridging pages 63-64 of the transcript Pumfrey J said this:

“In circumstances such as the present, I can see no reason why the familiar principles of American Cyanamid v Ethicon, with the necessary qualification that the question is no longer whether there is a triable issue, but whether there is a prospect of an arguable appeal should [not be] the principles to apply. And so in a case in which I thought there was a real prospect of success on an appeal, I would myself be inclined to continue the injunctive relief if it had been granted before trial.”

Pumfrey J then exercised the discretion concerning interim relief on the basis that there was no real prospect of success. He refused an injunction pending the outcome of the appeal from which he had given leave. He nevertheless granted a short injunction to enable the Court of Appeal to consider it. On appeal from his judgment, [2007] EWCA Civ 783, it was common ground that the normal American Cyanamid principles applied. Jacob LJ agreed that the appeal had no real prospect of success and therefore refused the injunction on that ground.

Pumfrey J’s view that the principles which apply to the grant of an interim injunction pending appeal are the same as those which apply to an injunction pending trial receives support from a decision of this court in Ketchum International v Group Public Relation Holdings [1997] 1 WLR 4. That case concerned the exercise of the Mareva jurisdiction, where of course the test applied is that of “good arguable case” rather than “triable issue”. The claimant had lost at the trial. Stuart Smith LJ said this at 10 G-H:

“Moreover, I cannot see any reason in principle why the considerations which are applicable when the court is considering the grant of a Mareva injunction should not be applied in favour of a plaintiff, even if he has lost in the court below, though the question will not be ‘Does he have a good arguable case?’ but ‘Does he have a good arguable appeal?’ This is likely to be a more difficult test to satisfy, and, if the case turns upon questions of fact which the judge has resolved against the plaintiff, may well be insuperable. This threshold must be at least as high as that which has to be satisfied when the court considers whether or not to grant leave to appeal where that is required.”

It is of course true, as Robert Walker LJ observed in Guardian Group v Associated Newspapers (unreported CA 20 January 2000) that the court, in applying the American Cyanamid principles, may give

“proper weight to any clear view which the court can form at the time of the application for interim relief (and without the need for a mini-trial on copious affidavit evidence) as to the likely outcome at trial” (see [18]).

Guardian Group v Associated Newspapers was, however, a passing off case. In assessing the balance of justice in those cases it is frequently necessary to form a view as to the strength of the plaintiff’s claim in order to understand the scale of any likely damage: see the judgment of Robert Walker LJ at [13] - [14]. In the more general run of case the courts have no corresponding need to delve into the merits in order to resolve the question of what interim protection is appropriate in any individual case.

In Virgin Atlantic Airways v Premium Aircraft Interiors [2009] EWCA Civ 1513 the Court of Appeal was concerned with the difference between an injunction pending a possible appeal to the Supreme Court and exceptions to an injunction where there is no possibility of appeal. In relation to the first of these, Jacob LJ cited at [21] the well known passage from the Minnesota Mining case which I have quoted above. He called this “the balance of convenience principle”. He then said at [22]:

“It should be noted the question is not the same when one is considering what to do on an application for an interim injunction pending trial. In that case the patentee has yet to establish his right, whereas after successful trial he has prima facie done just that. So in general, when an appeal is pending, the patentee will get his injunction provided he gives a cross-undertaking in damages against the possibility that the defendant's appeal would be successful. The question, however, remains one of a balance of convenience. ”

I do not think that in that passage from his extempore judgment in that case, Jacob LJ was intending to qualify anything said by Buckley LJ in Minnesota Mining. Following trial a successful patentee is entitled to a permanent injunction. Moreover it is obvious that, whatever the strength of the defendant’s appeal, the patentee has real prospects of maintaining that injunction at the conclusion of the appeal. Whether the injunction should be stayed, and the conditions on which a stay should be granted or refused are, as Jacob LJ recognised in the final sentence, dependent on the balance of hardships or convenience.

In Servier Pumfrey J also rejected a submission by counsel for the defendants that the only ground for continuing the relief would be that the appeal would be rendered nugatory if relief was refused. He accepted the submission of counsel for the claimant that the jurisdiction, which derives from section 37 of the Supreme Court (now Senior Courts) Act 1981 was a wide one.

I would summarise the principles which apply to the grant of an interim injunction pending appeal where the claimant has lost at first instance as follows:

With those principles in mind I turn to the first ground of appeal, which is that the judge should not have held that the decision on the merits affected the balance of convenience or hardships. The judge plainly did regard the balance of hardships prevailing pending the appeal as different from that which prevailed prior to judgment. In doing so, he gave weight to the fact that the merits had been decided at first instance and that it was not possible to say that the case was more than “plainly arguable”. In particular he said that the balance might have been tipped in Novartis’ favour if there were strong prospects of success.

Relying on Guardian Group v Associated Newspapers (supra) Mr Speck QC for Hospira submitted that the decision on the invalidity of the patents was a “clear view on the merits” which the court could properly take into account in exercising the discretion to refuse Novartis interim relief.

Dr Turner QC for Novartis takes as his primary position that the judgment of Arnold J does not affect the exercise of the discretion. If he is wrong about that he submits that there may be cases, such as those where a patentee has lost on a heavily fact dependent issue such as obviousness, where it may be proper to take the judge’s view into account, because in those cases there may be a heavy presumption that the judge’s view will prevail. He submits that the present case is different. Here success on appeal depends on the issue of priority, which is essentially a question of construction of documents.

In my judgment this is not a case where one can say that there is a “clear view on the merits” which ought to be taken into account in the exercise of discretion pending appeal. Both Arnold and Birss JJ formed the view that an appeal would have a realistic prospect of success. Having been shown the material, I share those views. I do not think that the exercise of discretion should be approached on the basis that the merits are clear in favour of one side or the other.

Mr Speck also supported the decision of the judge on the basis that depriving a successful litigant of the fruits of his success pending appeal was inherently unfair, and that the judge had therefore been right to put weight in the scales in favour of Hospira on this ground. This submission is linked to, and in my judgment contingent on the second ground of appeal, under which it is argued that there was a tacit assumption that the trial at first instance would have cleared Hospira to market their product.

I do not think that the approach adopted by the judge was correct. Once he was satisfied that the appeal had real prospects of success, he should, on the facts of the present case, have gone on to consider the balance of hardships in accordance with the principles I have set out above. He should not have put a weight into the scales against Novartis merely because they could not satisfy some higher threshold on the merits. Further, for reasons I explain in relation to the second ground, I do not think that this is a case where it could be said that there is substantial unfairness in declining to give Hospira the right to rely on their judgment pending appeal if the other considerations of unquantifiable loss to each side would otherwise justify an injunction.

The second ground of appeal is that the judge attached weight to the letter of 18 December 2012. There is no dispute that he did attach considerable weight to this letter. It was said to illustrate the fact that the trial was scheduled to take place before 15 May 2013 so as to clear the use patents out of the way and enable Hospira to launch its product as soon as the SPC expired. It had been argued before him that it was wrong to force generic medicine manufacturers to start litigation such as this so long before the anticipated launch date to enable an appeal to the Court of Appeal to take place beforehand. The judge said that if Novartis were going to argue for such a proposition they should have asked for the case to be scheduled sufficiently far in advance of 15 May 2013 to allow an appeal to the Court of Appeal to take place before that date. It appears that this was an important factor behind the judge’s conclusion that the risk of significant uncompensatable harm should fall on Novartis.

I have already explained that the February trial date was one which was set without recourse to any expedition. In order to cause the trial date to be scheduled sufficiently far in advance of May 2013 for an appeal to have been accommodated before that date, one of the parties would have had to apply to the court for expedition. I have great difficulty in understanding how it can be said that Novartis should have taken any steps to apply for expedition before early December 2012 when they became aware that Hospira had obtained a marketing authorisation. Until that date they would not have been able to say that there was any commercial urgency from their point of view. Even after that date they were unaware of what plans for launching a product Hospira actually had. Hospira made no firm statement of their intention to launch until after the trial.

Recognising these difficulties, Mr Speck suggests that Novartis should have raised the issue at an earlier date. I take this to mean that Novartis should have written to Hospira to notify them that if Novartis were unsuccessful at the trial they would nevertheless seek an injunction pending appeal, or at least that Hospira should not regard the proceedings as concluded in their favour until all Novartis’ possible avenues of appeal had been exhausted. This would then have enabled Hospira to make the necessary arrangements for an expedited trial.

I think it was wrong in principle for Novartis to be required to assume this burden. Both parties were aware that litigation is not finally concluded until all appeals are disposed of. It was put well by Cotton LJ in Polini v Gray (1879) 12 Ch. D 438 at 446when explaining the principle on which the court acts to stay an order it has made:

“It does so on this ground, that when there is an appeal about to be prosecuted the litigation is to be considered as not at an end...”

This proposition rings particularly true in the pharmaceutical industry. In Smithkline Beecham v Apotex [unreported CA 23 October 2001], Jacob LJ articulated the need in the pharmaceutical industry for a generic manufacturer who makes plans to launch a generic medicine, to take steps to clear the obstacles facing its manufacture out of the way before it is launched. He said

“You would have to be very naïve in the pharmaceutical industry to think that the patentee, with a product as important as this, would not, if it had anything other than a frivolous chance of success, take action.”

It was argued in that case that it was wrong in principle to place the burden on the generic company. The response to that was:

“I see no question of principle involved here of any sort. It is purely common sense. If there may be an obstacle in your way, clear it out. To my mind this is a case where the retention of the status quo is a rational thing to do. It was something that could have been avoided by the defendants. They chose not to do it.”

The way to market for a generic manufacturer is not clear until all arguable objections from the patentee have been eliminated. If the generic manufacturer allows the trial of the action at first instance to coincide with the intended launch date he runs the risk that a successful appeal could get in the way, even if judgment at first instance is given in his favour.

I think it is also wrong to suggest, if the judge was so suggesting, that the parties were proceeding to trial on the tacit understanding of both parties that the decision at first instance on the validity of the use patents would determine Hospira’s freedom to sell the product covered by their marketing authorisation after 15 May 2013. The letter of 18 December 2012 was, as I have explained, written to resolve doubts about whether, if Novartis were successful, Hospira would nevertheless maintain that they were free to launch their product because the issue of infringement had not been determined against them. It is unrealistic to read from this letter, or the other letters relied on by Hospira to support this suggestion, a tacit understanding that Novartis would not pursue any avenue open to them to restrain the launch of the product. Hospira’s solicitors’ letter of 10 April in which they stated that the enforceability of the patents by way of interim injunction depended on the outcome of Novartis’ application for permission to appeal was an accurate statement of the position.

I consider therefore that both grounds of appeal are made out. It does not follow that the injunction should be granted. For this reason we heard argument from both parties as to how the discretion should be exercised if we were to exercise it afresh.

Mr Turner submitted, in essence, that the judge had got it right when he considered that the facts of this case would have justified the grant of an injunction before the trial. He should not have allowed the scales to be tipped against Novartis by the two factors which I have already considered, that is to say the fact that judgment had been given against Novartis and the points made surrounding the letter of 18^th December 2012.

Mr Speck submitted that this was a case where an injunction should not be granted. So far as the damage to his clients was concerned, he stressed that this was a market which, because it proceeded mainly by regional hospital tendering contracts, could be penetrated rapidly and that Hospira were poised to do so, and thereby take advantage of a unique opportunity. Before the judge it was known that Hospira had been awarded the Scottish contract and it is now known that they have been awarded the Welsh contract as well. The judge had been right to say that there was some risk of Hospira losing a first mover advantage that it would otherwise have obtained. Moreover the impact of the injunction would be to restrain the sales of the authorised product for the indication of Paget’s disease, to which there could be no lawful objection.

So far as the loss to Novartis was concerned Mr Speck repeated many of the attacks on the likelihood of price cutting which were advanced before the judge. He emphasised that Novartis have a related company, Sandoz, who also had the benefit of marketing authorisation to sell a generic product. Thus Novartis could avoid loss of market share without price cutting their branded product, by making generic sales through Sandoz.

Mr Speck also stressed two public interest points:

Despite Mr Speck’s able submissions, I consider that this is a case where the court should grant an injunction pending the appeal to restrain the sale of the Hospira product. The judge was in my judgment plainly right that this was a case where, had Hospira sought to launch before trial, they would have been restrained until trial. The damage to Novartis was both more certain to occur and greater in magnitude than the damage to Hospira. Although there would be difficulties of knowing for certain whether Hospira had lost a first mover advantage, it does appear from the information given to us on this appeal that at least one and probably two other generic companies are in a position to market products the subject of authorisations. In those circumstances the possibility of the period between now and trial being one during which Hospira would have been able to enjoy the benefits of being a first mover seems rather more remote.

There was no precise evidence on the extent of the market for Paget’s disease. The judge concluded that it was very small. Whilst this is a factor, like the judge I do not think it plays much part in the overall assessment.

The unquantifiable damage to Novartis seems to me to outweigh that to Hospira. From the evidence, an immediate downward price spiral, even in the period between now and the hearing of the appeal, seems highly likely if not inevitable. The fact that Novartis can divert sales to Sandoz does not vitiate this conclusion. They will still be faced with the fact that the market will have become accustomed to lower prices, and restoring their monopoly position will, if possible at all, be accompanied by harm of other kinds.

I am not persuaded by either of the public interest points. As to the first point, appeals are the norm rather than the exception in this type of case. It is not unreasonable to expect generic companies to plan on the basis that there will be an appeal. To do so will not make the process of clearing the way any more expensive, as Mr Speck at one point suggested it might. It means only that generic companies will have to disclose their intentions to the innovator in time for an appeal to take place, or, if this is not possible, seek an appropriately expedited hearing to take this into account. As to the second point, it is true that the National Health Service will not achieve 100% recovery under the cross-undertaking which it has been offered. That is, of course, unfortunate, but is the case whenever the court grants an injunction to restrain sales of a pharmaceutical which turns out not to be justified. In the present case it is not a point of sufficient weight to justify refusing the injunction.

In my judgment the judge was right to say that the best way of holding the ring pending trial would have been to grant an injunction (had one been necessary) but wrong to say that the same did not apply pending the appeal. I would accordingly allow the appeal and grant the injunction until final judgment in the appeal or further order.

I agree.

I also agree.