Apotex Pty Ltd v Les Laboratoires Servier

IN THE FEDERAL COURT OF AUSTRALIA NEW SOUTH WALES DISTRICT REGISTRY GENERAL DIVISION NSD 51 of 2012

AND: LES LABORATOIRES SERVIER AND ANOTHER AS NAMED IN THE SCHEDULE

AND BETWEEN: LES LABORATOIRES SERVIER AND ANOTHER AS NAMED IN THE SCHEDULE

AND: APOTEX PTY LTD ACN 096 916 148 AND ANOTHER AS NAMED IN THE SCHEDULE

JUDGE: RARES J DATE OF ORDER: 12 MARCH 2015 WHERE MADE: SYDNEY

1. The applicant/cross-respondent and the opponent pay 60% of the respondent/cross-claimant's costs of the application to amend.
2. The applicant/cross-respondent pay 66.67% of the respondent/cross-claimant's costs of the proceedings. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

IN THE FEDERAL COURT OF AUSTRALIA NEW SOUTH WALES DISTRICT REGISTRY GENERAL DIVISION NSD 51 of 2012

AND: LES LABORATOIRES SERVIER AND ANOTHER AS NAMED IN THE SCHEDULE

and between: LES LABORATOIRES SERVIER AND ANOTHER AS NAMED IN THE SCHEDULE

AND: APOTEX PTY LTD ACN 096 916 148 AND ANOTHER AS NAMED IN THE SCHEDULE

JUDGE: RARES J DATE: 12 MARCH 2015 PLACE: SYDNEY

REASONS FOR JUDGMENT (REVISED FROM THE TRANSCRIPT) 1 The question for decision is what orders for costs should be made in these proceedings. The successful party has lost on nearly all of the issues it litigated during a three-week long trial and a subsequent one-week long application to amend a patent under s 105 of the Patents Act 1990 (Cth): Apotex Pty Ltd v Les Laboratoires Servier [2013] FCA 1426 (the trial reasons); Apotex Pty Ltd v Les Laboratoires Servier [2015] FCA 104 (the amendment reasons).

Issues 2 The parties raised a number of substantive questions on the issue of costs, namely: (1) whether in patent litigation, or litigation generally, the Court should exercise its undoubted discretion under s 43 of the Federal Court of Australia Act 1976 (Cth) and, in particular, under s 43(3)(e), to order costs against a party whether or not the party was successful in the proceeding; (2) whether, in the context of an amendment application under s 105 of the Patents Act, it is appropriate to depart from the general rule that the applicant patentee, seeking an amendment, ought pay the costs of any opposing party in the litigation as well as the costs of any third party opponent that exercises its right to intervene to oppose the amendment application, whatever the outcome on the issues raised by the opponents; (3) whether an offer of compromise under r 25.14(3) of the Federal Court Rules 2011 (Cth) made by Apotex ought create either a presumptive entitlement in its favour in respect of costs incurred before the making of the order on a party/party basis, and then from the two business days after it was served on an indemnity basis, or whether a corresponding offer made under the principles in Calderbank v Calderbank [1976] Fam 93 ought produce a similar result.

Principles 3 Relevantly, s 43 of the Federal Court of Australia Act provides: 43 Costs (1) The Court or a Judge has jurisdiction to award costs in all proceedings before the Court (including proceedings dismissed for want of jurisdiction) other than proceedings in respect of which this or any other Act provides that costs must not be awarded. This is subject to: (a) subsection (1A); and (b) section 570 of the Fair Work Act 2009; and (c) section 18 of the Public Interest Disclosure Act 2013. … (2) Except as provided by any other Act, the award of costs is in the discretion of the Court or Judge. (3) Without limiting the discretion of the Court or a Judge in relation to costs, the Court or Judge may do any of the following: … (b) make different awards of costs in relation to different parts of the proceeding; (c) order the parties to bear costs in specified proportions; (d) award a party costs in a specified sum; (e) award costs in favour of or against a party whether or not the party is successful in the proceeding; … (g) order that costs awarded against a party are to be assessed on an indemnity basis or otherwise; … Note: For further provision about the award of costs, see subsections 37N(4) and (5) and paragraphs 37P(6)(d) and (e). (emphasis added) None of the exceptions in s 43(1) applies here. 4 Rule 25.14(3) provides: 25.14 Costs where offer not accepted (3) If an offer is made by an applicant and not accepted by a respondent, and the applicant obtains a judgment that is more favourable than the terms of the offer, the applicant is entitled to an order that the respondent pay the applicant's costs: (a) before 11.00 am on the second business day after the offer was served - on a party and party basis; and (b) after the time mentioned in paragraph (a) - on an indemnity basis. Note 1: Costs on an indemnity basis is defined in the Dictionary. Note 2: The Court may make an order inconsistent with these rules - see rule 1.35. (emphasis added)

Principles as to costs 5 It is necessary to say something about the Court's power to order costs before proceeding to deal with the arguments of the parties, some of which appeared to ignore the fact that this power consisted of a judicial discretion. The ordinary principles applied by the Court in considering an award of costs were neatly stated by French CJ, Hayne, Bell, Gageler and Keane JJ in Gray v Richards (No. 2) (2014) 89 ALJR 113 at 113-114 [2] as follows: The disposition of costs is within the general discretion of the Court. Ordinarily, that discretion will be exercised so that costs are awarded to the successful party, but other factors may have a significant claim on the discretion of the Court [Stewart v Atco Controls Pty Ltd (In liq) (No 2) (2014) 252 CLR 331 at 334 [4]]. The disposition which is ultimately to be made in any case where there are competing considerations will reflect a broad evaluative judgment of what justice requires. (emphasis added) 6 Earlier, Gleeson CJ, Gummow, Hayne and Crennan JJ had said in Foots v Southern Cross Mine Management Pty Limited (2007) 234 CLR 52 at 62-63 [25]-[26] with respect to a statutory power similar to that in s 43(1) and (2) of the Federal Court of Australia Act: Before considering those submissions, several general propositions regarding an award of costs should be noted. First, the award is discretionary but generally that discretion is exercised in favour of the successful party; in Queensland, this general understanding is expressed in r 689(1) of the UCPR, which states: "[c]osts of a proceeding, including an application in a proceeding, are in the discretion of the court but follow the event, unless the court considers another order is more appropriate." Further, although capable of estimation, the actual monetary value of an award of costs cannot be ascertained until those costs are taxed or otherwise assessed. In Oshlack v Richmond River Council [(1998) 193 CLR 72], this Court emphasised the breadth of the power exercised in making costs orders. There the trial judge had made no order as to costs despite the Council's success, and this decision was affirmed on appeal to this Court. Gaudron and Gummow JJ rejected a submission that there was an automatic rule that costs always follow the event [(1998) 193 CLR 72 at 88 [40]-[41] (footnote omitted)]: "There is no absolute rule with respect to the exercise of the power conferred by a provision such as s 69 of the [Land and Environment] Court Act [1979 (NSW)] that, in the absence of disentitling conduct, a successful party is to be compensated by the unsuccessful party. Nor is there any rule that there is no jurisdiction to order a successful party to bear the costs of the unsuccessful party. If regard be had to the myriad circumstances presenting themselves in the institution and conduct of litigation, and to the varied nature of litigation, particularly in the equity jurisdiction, it will be seen that there is nothing remarkable in the above propositions." (emphasis added) 7 Nonetheless, the principle is that costs orders are not made to punish an unsuccessful party: Foots 234 CLR at 63 [27], endorsing what Gaudron and Gummow JJ had said in Oshlack v Richmond River Council (1998) 193 CLR 72 at 88-89 [44]. 8 The true position under the Judiciary Act 1903 (Cth) principles found in provisions such as s 43 of the Federal Court of Australia Act is that encapsulated in the citation by the majority in Foots at 234 CLR at 65 [34] who endorsed a passage from Daniell's Practice of the High Court of Chancery (5th ed (1871) Vol 2, p 1239), that stated: In exercising this discretion, however, the Court does not consider the costs as a penalty or punishment; but merely as a necessary consequence of a party having created a litigation in which he has failed … 9 Those principles were applied by Greenwood and Rares JJ in Kazar v Kargarian (2011) 197 FCR 113 at 115-117 [2]-[9]. There is no need for a successful party to have engaged in some form of delinquent conduct so as to actuate the power of the Court to make an order for costs against it. These principles apply in patent litigation, as has been clear since at least the decision of the Full Court in Probiotec Limited v University of Melbourne (2008) 166 FCR 30, where I discussed the principles at 42-44 [45]-[50] with the agreement of Finn and Besanko JJ. 10 In a number of earlier decisions in patent cases, single judges of the court have ordered that the amount of costs awarded to the successful party be reduced, although none made an order that the successful party should bear a portion of the unsuccessful party's costs. 11 In Patent Gesellschaft AG v Saudi Livestock Transport and Trading Company (1996) 33 IPR 461 at 465 Olney J made an order that each party pay its own costs of the proceedings rather than making the unsuccessful party pay costs where a respondent had chosen, for tactical reasons, to plead at enormous length many particulars of invalidity that it did not ultimately press and it had also caused the unnecessary prolongation of the trial. There his Honour found each party had had a measure of success on an issue by issue basis. 12 In Gambro Pty Limited v Fresenius Medical Care South East Asia Pty Limited [2000] FCA 407 at [3] Tamberlin J ordered that the unsuccessful patentee on an application to amend a patent pay 70% of the opponent's costs. There, the successful party had made a long and sustained attack relating to matters of nondisclosure and the maintenance of a disputed claim in a patent. His Honour's reasoning was referred to without disapproval by Bennett J in Apotex Pty Limited v Les Laboratoires Servier (No 3) [2009] FCA 1069 at [5]. 13 Heerey J discussed the jurisdiction to make orders, including orders that the successful party pay the unsuccessful party's costs in patent litigation, in Eli Lilly and Co v Pfizer Research and Development Co NV/SA (2003) 59 IPR 234 at 238 [25] where he said correctly: As to costs, it is said that, according to Terrell on the Law of Patents, 14th ed, Sweet & Maxwell, London, 1994, p 245, the normal practice is that the applicant for an amendment pays the respondent's costs: see also Wimmera Industrial Minerals Pty Ltd v RGC Mineral Sands Ltd [1997] FCA 1337. However, there must remain a general discretion. Patent litigation, especially over pharmaceutical patents, is quite expensive enough, with the costs being passed on to consumers in the form of higher prices. I do not think it desirable that parties be encouraged to pursue unmeritorious objections like those in the present case by a belief that they will get their costs paid in any event. Because of the public interest element, applications to amend, at least where the amendment is by deletion, are not truly comparable to the "indulgence" said to be involved in the amendment of pleadings or the grant of adjournments. Indeed, there is some ground for thinking that Lilly and Bayer ought to pay Pfizer's costs, but since the latter moderately sought an order only for costs in the cause I will make an order accordingly. (emphasis added) 14 Lee J made similar observations in Innovative Agricultural Products Pty Limited v Crawshaw (No 2) [1996] AIPC 91-281 at p 37,829. I agree with those observations and with what Nicholls LJ, with whom Fox LJ and Sir David Croom-Johnson concurred, said in C Van Der Lely NV v Ruston's Engineering Co Ltd [1993] RPC 45 at 63 namely that: in appropriate cases, the costs order at the trial should reflect the extent to which significant sums of costs have been thrown away by reason of one party, albeit successful overall, raising and pursuing unsuccessful points.

Background 15 The issues both in the trial and the subsequent amendment proceedings are set out in the trial reasons and the amendment reasons. In substance, Apotex won on a discrete, single issue at a trial in which it impugned and litigated the validity of Servier's patent on many fronts. That winning issue was the best-method issue. In all other substantive respects, Servier's position was vindicated, on challenges involving the lengthy and time consuming issues of inventive step, obviousness and false suggestion. During the interlocutory stages and in the course of the trial, Apotex abandoned its challenges to the utility of the patent and, after seeing the experts' joint report, to the sufficiency of the description in the patent. 16 Moreover, Apotex chose to commence the litigation, as its submissions on the costs application explained, to "clear the way". That is, when it began the proceedings in January 2012, Apotex engaged in a considered attack on the validity of the patent, which it was able to prepare at leisure and on which it struck when it considered that it was ready to pursue the litigation to maximum effect for its commercial benefit. That was its right. However, in exercising that right, Apotex chose the battle grounds on which to assert that the patent was invalid at a time when, commercially, it was ready to launch its generic product. 17 Apotex accepted in the early stages of the proceedings, that, if the patent were valid, the conduct in which it proposed to engage would infringe. That led to the interlocutory injunction preventing Apotex from going to market so as to hold the status quo. That injunction was supported by Servier's undertaking as to damages on which it now is liable, having regard to the ultimate outcome reflected in the orders made on 26 February 2015 that revoked the patent on the basis of Apotex' success on the best method issue.

The offers of compromise 18 The trial commenced on 8 October 2013. Shortly before that, on 20 September 2013, Apotex made two offers to Servier to compromise the proceedings. The first was made under r 25.14(3), and the second was made as a Calderbank offer. The substance of each offer was the same. Apotex' solicitors sent the two offers with letters that commenced by asserting that Apotex was in an "extremely strong position", based on their review of the various affidavits that had been filed on both sides and having regard to the imminent trial. The letters went on to claim that Apotex should win not only on best method point but also on each of the points on which it ultimately failed. The letters and the formal notice of offer to compromise under r 25.01(1) made the following significant proposals as part of the cognate offers: Apotex' application for revocation be dismissed; Servier's claim for infringement also be dismissed; the interlocutory injunction be discharged and Servier be released from its undertaking as to damages; each party bear its own costs; and Servier grant Apotex a royalty-free, non-exclusive licence under, and for the remainder of the term of, the patent, including any applicable extension in terms substantially as set out in the offers. 19 In other words, both offers proposed that the litigation be completely resolved with neither party exposed to any financial obligations to the other, including for costs and in respect of Servier's undertaking as to damages, and that, significantly, Servier would grant Apotex a non-exclusive licence for the term of the patent. 20 Servier's solicitors responded on 23 September 2013 rejecting each offer and making a counter-offer. It is not necessary to detail the terms of the counter-offer.

Consideration - costs of the trial 21 Substantially, the whole of the time at the trial and the preparatory steps before trial were devoted to issues other than the best method issue. 22 Doing the best I can to arrive at a fair apportionment of the way in which Apotex and Servier ought bear costs of the proceedings prior to the bringing of Servier's amendment application, it seems to me that somewhere around about two days or so of the trial, and a similar proportion of the pre-trial, costs may have been relevant to the litigation of the best method issue. 23 Apotex' principal argument on that issue was available on the face of the patent, namely that it did not, as I found, provide anything more than a very general description of a classical method of salification. That description did not indicate any detail of the method that Servier had used to produce the salt, perindopril arginine, that was the active pharmaceutical ingredient in the tablets that were the subject of the test results set out in the complete specification. Nonetheless, Apotex was entitled to pursue the best method issue further and have discovery on it, which it obtained to the extent allowed. In the trial, Apotex ran that issue in a sensible and confined way, that I described in my earlier judgments. 24 Parties are entitled to choose the issues that they wish to litigate and to fight their cases on them. The Court's duty is, ordinarily, to decide those matters. But where parties choose arguments that can be segregated from others and where those arguments are wholly unsuccessful, there may be occasions, and patent cases are classic examples of them, where it is possible to make an appropriate apportionment to ensure that justice is done between the parties in the award of costs. The order at which I have arrived is not intended to express any form of punishment. It is intended to express what appears to me to be the only just way in which to deal with the fact that Servier has had to fight a lengthy, complex and expensive case that it won on many issues, but which ultimately resulted in its loss of its valuable patent. Servier lost on a very narrow point that need not have engaged the Court in the other substantial issues, that ultimately fell to be decided in its favour. 25 Subject to what I have to say about the offers of compromise, I think the justice of the case would be served by ordering that Apotex pay two thirds of Servier's costs of the litigation. That would reflect the substantial success that Servier enjoyed on all of the issues on which it ought not to have had to defend its patent in the litigation and on which it substantively succeeded: Gray 89 ALJR at 113-114 [2]; Foots 234 CLR at 62-63 [25]-[27]; C Van Der Lely [1993] RPC at 63; Eli Lilly 59 IPR at 238 [25].

Consideration - Costs of the amendment application 26 The ordinary rule in proceedings for an amendment to a patent is that the patentee pay the costs of the other parties to the proceedings and any opponent seeking to oppose the making of the amendment, regardless of the outcome. The justification for that rule of practice is not far to seek. The occasion for the amendment is brought about by an error that the patentee seeks to correct in respect of its earlier choice in framing the wording of its patent, that identified its claimed monopoly. 27 The question whether such a correction ought be made and its precise terms are matters of public interest, having regard to the monopoly rights conferred by a patent. Others who are, or potentially may be, affected by a proposed amendment, including the Commissioner of Patents, will be able to make any appropriate submissions and give appropriate assistance to the Court in its consideration of the disposition of the amendment application. That is particularly so, as I pointed out in the amendment reasons, that amendment applications can only be decided by the Court in proceedings in which the validity or enforcement of the patent is in issue ([2015] FCA 104 at [133]-[136]). Ordinarily, amendments are sought before, rather than after, the trial but, as I explained in those reasons, Servier's departure from the ordinary course in the present circumstance should not affect the result ([2015] FCA 104 at [171]-[179]). 28 Actavis argued that, in effect, there was an encrusted, and almost unbreakable, rule of law that whatever arguments an opponent to an amendment application might put up, it was entitled to do so at the cost of the patentee. Such an argument must be rejected. Costs are in the discretion of the Court, and they must always remain so under s 43 of the Federal Court of Australia Act. Moreover, if the position were as Actavis argued, great injustice could be done to a patentee by having to meet a barrage of unmeritorious, time wasting, or substantively unsuccessful arguments that need never have been run, simply because an opponent chose to exercise the supposed right of opposition knowing that whatever it said, its costs would be paid by the patentee. 29 It would be productive of injustice to fetter the discretion under s 43 to make orders for costs in the way for which Actavis contended. The discretion exists to enable the Court to make orders that are appropriate in all the circumstances, to do justice between the parties. 30 The ordinary rule for the patentee to pay the other parties' costs on an amendment application has a principled foundation to it, but in this case the opponents substantively lost on a large range of issues that they ran. I decided the amendment application substantively on the evidence that Servier initially filed in support of its application. That evidence met what I found to be its duty of full and frank disclosure of its reasons for seeking the amendment. However, as I explained in the amendment reasons, that duty was far more limited than the boundless duty advocated by the opponents during the course of the interlocutory stages and final hearing of the amendment application. The stance taken by the opponents on that issue, namely, the width of Servier's obligation of disclosure, caused the proceedings on the amendment application to be far more protracted than they ought to have been. 31 Throughout the course of the amendment application, from March 2014 when it was filed, Servier consistently argued, correctly, that its duty of disclosure consisted in having to explain fully and frankly why it was seeking the amendment and did not extend further, contrary to the opponents argument. The opponents argued, also from an early stage in the amendment application proceedings, that Servier had to disclose everything it knew about the best method at the time the complete specification was filed, including what it knew about methods other than 1986 and 1991 methods: see Apotex Pty Ltd v Les Laboratoires Servier (No 2) [2014] FCA 751 at [2], [7], [9] and [10]. I delivered those reasons immediately after the argument on 26 May 2014. 32 Ultimately, the correctness of the opponents' contention as to the extent of Servier's obligation of disclosure needed to be determined on the amendment application. Accordingly, I partially granted the opponents' applications for notices to produce to be issued to Servier, so that each of the parties could lead such evidence at the hearing of the amendment application as they wished, to make good their respective contentions: see Apotex Pty Ltd v Les Laboratoires Servier (No 3) [2014] FCA 1029 at [3], [7], [9] and [11]-[15]. 33 I have taken into account, in assessing the overall burden that any costs order might impose, the fact that, until the dying stages of the argument on 26 May 2014, Servier resisted adding to the declarations that it asked to be made, the last sentence in each identifying the fact that the 1986 and 1991 methods produced a white crystalline product. 34 A large amount of time was taken up during the course of the amendment application by the opponents' unsuccessful arguments that, first, the construction of s 105 of the Patents Act had a meaning other than the natural and ordinary meaning reflected in its clear words and that there was no power to grant the amendment sought to add the descriptions of the 1986 and 1991 methods and, secondly, that French and Lindgren JJ had contradicted themselves within 11 paragraphs in their careful reasons in Pfizer Overseas Pharmaceuticals v Eli Liley & Co (2005) 68 IPR 1 at 78 [379] and 80 [390]-[391]. Neither argument had any substance: see amendment reasons [2015] FCA 104 at [137], [143]-[145]. 35 Moreover, a large amount of the time and effort was taken up in the amendment application with the opponents' assertions that the 2002 method was, and was known by Servier to be, better than the 1986 and 1991 methods, even though there was no evidence at all that Servier knew at any time that the 2002 method was better (see amendment reasons: [2015] FCA 104 at [57]-[59], [69]-[72], [109]-[116]). The opponents also consumed time with another issue about their assertion of similarity between the 2002 method and the method in the alpha crystalline patent that required further expert evidence to be led. 36 Had I not allowed the notices to produce to be issued which, having regard to my ultimate finding as to Servier's obligation to give full and frank disclosure, I should not have, that material would not have been available to the opponents and the proceedings would have been very significantly shorter (cf: amendment reasons [2015] FCA 104 at [59]). 37 In effect, the opponents were seeking to use what I found was the wrongly asserted wide obligation of a patentee to make full and frank disclosure to force Servier to disclose anything that the opponents might then seize upon to suggest that it knew more than it had disclosed on the patent amendment application. In my opinion, the onus was on the opponents to propound and prove such a case, and before pleading it, they ought to have had a basis to do so. I was unable to see that basis, absent the material that was deployed, although that was not an issue before me, and I make no finding about it. However, it is necessary to refer to those matters in order to demonstrate that the opponents' erroneous reliance on an unduly wide extent of the patentee's disclosure obligation consumed a substantial amount of time and effort, not to say costs, in the conduct of the interlocutory and hearing stages of the amendment application. 38 During the course of argument today, the opponents persuaded me that I may have overstated the time that that hearing should have taken. I observed that it ought to have taken one day to hear (amendment reasons [2015] FCA 104 104 at [185]). However, I did not fully allow for the facts that Servier had initially deployed Mr Caine's evidence in support of its case for an amendment and that, when Dr Jaguelin had to give evidence, she did so through an interpreter which would have added somewhat to the time involved. Taking these matters into account, the hearing may have taken about one and half days, had it been confined to the real but limited issues. 39 Actavis argued in relation to its obligation to pay costs that Servier had somehow brought upon itself a liability to pay costs because, on 8 October 2014, it had added to the amendment application a claim that the 2002 method should be inserted into the complete specification. 40 I reject that argument. In the amendment reasons, I found that Servier made that application as a result of a contention in both opponents' amended grounds of opposition filed on 26 and 29 September 2014 ([2015] FCA 104 at [6]). That occurred after I had granted the opponents' application to serve limited notices to produce. The opponents' amended grounds of opposition raised the argument that the 2002 method was a better method known to Servier at the time that the complete specification was filed. 41 I also found in the amendment reasons that the parties did not raise any serious challenge to Dr Jaguelin's credibility and that she was generally honest and reliable. In making my finding ([2015] FCA 104 at [6]), I had in mind what Dr Jaguelin had said in her affidavit affirmed on 7 October 2014, namely that Servier was applying to add the 2002 method to its application to amend so as to avoid any doubt or substantive dispute that its proposed amendments did not include the best method of performing the invention known to Servier at the filing date. And, she said in the same paragraph of that affidavit, that the method used to produce the batch of perindopril arginine was the same as the description of the 2002 method the subject of the proposed further amendment. She said that the testing of that batch had not been completed or documented before the filing date. I accepted that evidence. 42 My findings in the amendment reasons reflected the fact that Dr Jaguelin's explanation demonstrated why Servier could not have known, whatever its state of belief was, that the 2002 method was better than either the 1986 or 1991 method. 43 Doing the best I can in all of the circumstances, and subject to what I have to say about the offers of compromise, I consider that justice would be done by ordering the opponents pay 60% of Servier's costs of the amendment application.

The offers of compromise - consideration 44 Two substantive issues arise, depending on the criteria to be applied in assessing each offer of compromise. The first issue is whether, under r 25.14(3), Apotex has obtained "a judgment that is more favourable than the terms of the offer". The second issue is whether, under the principles applicable to Calderbank offers, Servier has acted so as to attract a liability to pay indemnity costs by rejecting Apotex' offer. 45 In Apotex Pty Limited v AstraZeneca AB (No 5) [2013] FCA 560 at [10]-[11], Jagot J discussed a slightly different position, where Apotex had not made clear what the terms of a proposed licence would be and had not appeared to give up its rights under the undertaking as to damages. 46 In Novozymes A/S v Danisco A/S (No 2) [2013] FCAFC 55 at [12]-[15] Greenwood, Jessup and Yates JJ discussed an evolved form of an offer in the nature of that considered by Jagot J in AstraZeneca [2013] FCA 560. In that appeal, the party challenging the patent had offered to resolve the proceeding below on a similar basis to the present offer, namely, that it be granted a nonexclusive licence and that the challenge and enforcement proceedings be terminated with no orders as to costs. However, the issue there was whether acceptance of the offer would have resulted in a better outcome for the patentee under the principles applicable to a Calderbank offer. That decision does not assist on the issue arising here under r 25.14(3). 47 In my opinion, it is not possible to ascertain, on the material before me, whether the orders revoking the patent and (subject to the resolution of the present issue) for Apotex to pay costs in respect of the trial and amendment application amount to a judgment more favourable than the outcome Apotex proposed in its offer of compromise under r 25.14(3). Had Servier accepted that offer, there would have been no trial and no further exposure of the parties to their own solicitor/client obligations to pay their own costs or risk of being ordered to pay the other party's costs. Nor would Servier have been exposed on its undertaking as to damages for three years' worth of Apotex' lost profits. 48 The financial value of Servier's liability on its undertaking as to damages is, however, not susceptible of evaluation on the material before me: cf Waterside Workers' Federation of Australia v Stewart (1919) 27 CLR 119 at 132 where Isaacs and Rich JJ said, in relation to whether a performance bond could be characterised as a penalty, that "no one can ever tell how much loss is sustained by not doing business". That reasoning was approved by French CJ, Gummow, Crennan, Kiefel and Bell JJ in Andrews v Australia and New Zealand Banking Group Ltd (2012) 247 CLR 205 at 217 [11] in the context of whether the doctrine of penalties applied in relation to the selection of a fixed sum that would be due on a breach of contract where the prejudice to the interests of a party to whom that sum would be due was insusceptible of evaluation and assessment in money terms. Clearly, an undertaking as to damages is capable of enforcement but, ordinarily, that is done on evidence so as to quantify any loss or damage suffered by the person wrongly enjoined, whether or not such a sum would otherwise be recoverable under any legal or equitable cause of action. 49 The problem here, however, is that there is no evidence of what damages Apotex may be entitled to claim for the three years that it has been restrained from marketing its generic product. There is no basis to calculate what market price Apotex would have charged using the proposed licence, what sales and profit it would have made on that price or what, if any, competition, apart from that offered by Servier, it would have faced. Nor is there any evidence about what Apotex' profits might be over the balance of the term of the revoked patent, especially when other generic producers, such as Actavis, will be in the market. 50 Thus, there is no evidence on which to compare any difference in Apotex' financial position had its offer been accepted and its actual position now. Those two positions were capable of proof by Apotex to demonstrate the asserted superiority of the final outcome over that proposed in its offer under r 25.14(3). The absence of such evidence means that it is not possible to assess whether Apotex would have been better off had Servier accepted the offer, than the position that Apotex now has. 51 The automatic consequence of a costs order in favour of the applicant under r 25.14(3) depends upon the existence of the fact that the Court's judgment is actually more favourable to the applicant (here, Apotex) than the terms of its offer. 52 As a matter of common sense, if Servier had granted Apotex the licence under the offer it would have created a duopoly, in place of its existing monopoly, under the patent. There would be no apparent commercial reason for Servier to grant licences to other generic manufacturers and little likely incentive for them to produce under such licences, given that Servier would be able to charge a royalty. There is no self-evident reason why Servier would wish to let other competitors, beyond Apotex, into the market where the likely consequence of doing so would be to reduce both Servier's market share and profitability. These matters may or may not have been established by evidence, to demonstrate that Apotex' position now that the patent has been revoked and it is entitled to prove its damages but is liable to Servier for costs is more favourable to it than if Servier had accepted the r 25.14(3) offer in September 2013. 53 Had Servier accepted that offer, Apotex would have had the right to exploit the nonexclusive licence. However, I have no information at all as to what its financial position would have been in that circumstance, and whether that would have produced a result more or less favourable than what it now has. Accordingly, I am unable to make any finding on that critical matter. 54 For these reasons, I am not satisfied that Apotex has obtained a judgment that is more favourable than the terms of its offer, so as to engage the operation of r 25.14(3). 55 That brings me to the question of whether I should make an order in favour of Apotex in respect of its without prejudice Calderbank offer. 56 In order to establish that a party is entitled to the benefit of an order that its opponent pay indemnity costs, based on without prejudice costs offers outside the offer of compromise regime in the Rules, it is necessary for the party seeking indemnity costs to demonstrate that the other party's refusal of the Calderbank offer was unreasonable: see Elecspess Pty Limited v LED Technologies Pty Limited (2013) 215 FCR 95 at 101 [16] per Besanko and Jessup JJ; Insight SRC IP Holdings Pty Limited v Australian Council for Education Research Limited (No 2) [2013] FCAFC 73 at [8] per North, Rares and Robertson JJ; Sagacious Legal Pty Limited v Wesfarmers General Insurance Limited [2011] FCAFC 53 at [128]-[132] per Besanko, Perram and Katzmann JJ. 57 In my opinion, Apotex has not demonstrated that Servier's refusal of the Calderbank offer was unreasonable. Each party clearly perceived its own case in the litigation at the time the offer was made to be strong. Each was vindicated in its contemporaneous assessment of the strength of parts of its case to some degree. But, in Apotex' case, the end result came nowhere near the degree of success that its solicitors' letter of offer suggested. The effect of Apotex' offer would have been to leave the validity of the patent, to some extent, uncertain and to open up the field for Apotex, to the limited degree of the nonexclusive licence. 58 Moreover, Servier's senior counsel's advice on the best method issue on which, ultimately, Servier lost, had been given to its solicitors on about 17 July 2013: amendment reasons [2015] FCA 104 at [24]. That advice was to the effect that Servier was not at risk on the best method issue. I am of opinion that the advice provided reasonable grounds for Servier to consider that it would be able to meet the best method issue successfully, together with all of the other challenges to the validity of the patent. 59 I cannot see that the fact that the two sides genuinely held different views, as to their respective prospects of success, rendered Servier's refusal to accept Apotex' offer unreasonable or, "imprudent or plainly unreasonable", to use the expression of Besanko and Jessup JJ in Elecspess at 215 FCR 101 [16]. 60 In my opinion, that description would have been unfair as a description of the basis for any perception of the strength of the cases in the litigation of each of the parties at the time that Apotex' offer was made. 61 Accordingly, I am not satisfied that the orders that I have suggested for costs ought be affected by the offers of compromise, on either basis on which Apotex relied.

Conclusion 62 For these reasons, I will order that: (1) Apotex and Actavis pay 60% of Servier's costs of the application to amend. (2) Apotex pay 66.67% of Servier's costs of the proceedings. I certify that the preceding sixty-two (62) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Rares.

Associate: Dated: 13 April 2015 SCHEDULE NSD 51 of 2012 LES LABORATOIRES SERVIER First Respondent SERVIER LABORATORIES (AUST) PTY LTD Second Respondent AND BETWEEN: LES LABORATOIRES SERVIER First Cross-Claimant SERVIER LABORATORIES (AUST) PTY LTD Second Cross-Claimant APOTEX PTY LTD ACN 096 916 148 First Cross-Respondent SYMBION PTY LTD ACN 000 875 034 Second Cross-Respondent