H. Lundbeck A/S v Sandoz Pty Ltd

Date of hearing: 16, 17, 18, 19, 24, 26 and 30 April 2018, 1, 2, 3 and 4 May 2018, 2, 3, 4, 5, 9 and 10 October 2018

Date of last submissions: 10 October 2018

Registry: New South Wales

Division: General Division

National Practice Area: Intellectual Property

Sub-area: Patents and associated Statutes

Category: Catchwords

Number of paragraphs: 551

Counsel for H. Lundbeck A/S, Lundbeck Australia Pty Ltd and CNS Pharma Pty Ltd: AJL Bannon SC with KJ Howard SC, LA Merrick and C Cunliffe

Solicitor for H. Lundbeck A/S, Lundbeck Australia Pty Ltd and CNS Pharma Pty Ltd: Corrs Chambers Westgarth

Counsel for Alphapharm Pty Ltd, Apotex Pty Ltd and Aspen Pharma Pty Ltd: DKC Catterns QC with S Habib SC, C Dimitriadis SC, N Murray SC, AR Lang and B Mee

Solicitor for Alphapharm Pty Ltd, Apotex Pty Ltd and Aspen Pharma Pty Ltd: King & Wood Mallesons

Counsel for Sandoz Pty Ltd: JM Hennessy SC with PL Arcus and J Adamopoulos

Solicitor for Sandoz Pty Ltd: Clayton Utz

BETWEEN: H. LUNDBECK A/S First Applicant

LUNDBECK AUSTRALIA PTY LTD (ACN 070 094 290) Second Applicant

AND: SANDOZ PTY LTD (ACN 075 449 553) Respondent

JUDGE: JAGOT J DATE OF ORDER: 21 November 2018

1. Subject to order 2, until 5.00pm on 28 November 2018, pursuant to s 37AF of the Federal Court of Australia Act 1976 (Cth) and on the ground that it is necessary to prevent prejudice to the proper administration of justice under s 37AG(1)(a), there be no disclosure (by publication or otherwise) of the reasons for judgment delivered on the date of this order in proceedings NSD 647 of 2014 and NSD 826 of 2016 (Proceedings) to any person other than to the external solicitors, expert accounting witnesses, and counsel for the parties in the Proceedings and Lundbeck In-house Counsel and Sandoz In-house Counsel (as defined in the Confidentiality Agreement dated 30 May 2017).
2. The reasons for judgment may be provided to the Patent Office in relation to the Application for Licence to Exploit an Invention made by the Respondent to the Patent Office on 18 December 2013, provided that the Commissioner of Patents makes a direction under regulation 4.3(2)(a) and / or 4.3(2)(b) that the reasons for judgment will not be open to public inspection pending the making of any confidentiality orders referred to in order 6.
3. By 4.00pm on 28 November 2018 any party wishing to claim that any part of the reasons for the judgment should be subject to a further confidentiality order is to notify the Associate to Jagot J and the other parties, by email of the claim including: (a) details of the matter claimed to be confidential; (b) a short statement of the reasons the matter is said to be confidential; and (c) a statement identifying whether the claimant consents to the confidentiality claim being determined by Jagot J on the basis of the email or seeks an oral hearing.
4. If no notice by email is received in accordance with order 3, the reasons for judgment will be published forthwith.
5. If notice is received in accordance with order 3, the reasons for judgment will be published forthwith with the claimed confidential matter redacted pending determination of the confidentiality claim.
6. The parties are to confer and, by 4.00pm on 5 December 2018, are to propose in a joint email (including agreed and disagreed matters) to the Associate to Jagot J further directions to enable the matter to be finalised, including orders to enable the expert accounting witnesses to confer and provide a further joint report, if necessary, in respect of any outstanding issues having regard to the reasons for judgment.
7. Pursuant to rule 36.03(b) of the Federal Court Rules 2011 (Cth), the time within which a party may file and serve any notice of appeal from judgment in these proceedings be extended to and for a period of 21 days after the publication of the reasons for judgment pursuant to orders 4 or 5, or the date of the determination of any outstanding issues raised by the expert accounting witnesses in any further joint report pursuant to order 6, whichever is later. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

BETWEEN: CNS PHARMA PTY LTD (ACN 121 515 400) Applicant

AND: SANDOZ PTY LTD (ACN 075 449 553) Respondent

JUDGE: JAGOT J DATE OF ORDER: 21 November 2018

1. Subject to order 2, until 5.00pm on 28 November 2018, pursuant to s 37AF of the Federal Court of Australia Act 1976 (Cth) and on the ground that it is necessary to prevent prejudice to the proper administration of justice under s 37AG(1)(a), there be no disclosure (by publication or otherwise) of the reasons for judgment delivered on the date of this order in proceedings NSD 647 of 2014 and NSD 826 of 2016 (Proceedings) to any person other than to the external solicitors, expert accounting witnesses, and counsel for the parties in the Proceedings and Lundbeck In-house Counsel and Sandoz In-house Counsel (as defined in the Confidentiality Agreement dated 30 May 2017).
2. The reasons for judgment may be provided to the Patent Office in relation to the Application for Licence to Exploit an Invention made by the Respondent to the Patent Office on 18 December 2013, provided that the Commissioner of Patents makes a direction under regulation 4.3(2)(a) and / or 4.3(2)(b) that the reasons for judgment will not be open to public inspection pending the making of any confidentiality orders referred to in order 6.
3. By 4.00pm on 28 November 2018 any party wishing to claim that any part of the reasons for the judgment should be subject to a further confidentiality order is to notify the Associate to Jagot J and the other parties, by email of the claim including: (a) details of the matter claimed to be confidential; (b) a short statement of the reasons the matter is said to be confidential; and (c) a statement identifying whether the claimant consents to the confidentiality claim being determined by Jagot J on the basis of the email or seeks an oral hearing.
4. If no notice by email is received in accordance with order 3, the reasons for judgment will be published forthwith.
5. If notice is received in accordance with order 3, the reasons for judgment will be published forthwith with the claimed confidential matter redacted pending determination of the confidentiality claim.
6. The parties are to confer and, by 4.00pm on 5 December 2018, are to propose in a joint email (including agreed and disagreed matters) to the Associate to Jagot J further directions to enable the matter to be finalised, including orders to enable the expert accounting witnesses to confer and provide a further joint report, if necessary, in respect of any outstanding issues having regard to the reasons for judgment.
7. Pursuant to rule 36.03(b) of the Federal Court Rules 2011 (Cth), the time within which a party may file and serve any notice of appeal from judgment in these proceedings be extended to and for a period of 21 days after the publication of the reasons for judgment pursuant to orders 4 or 5, or the date of the determination of any outstanding issues raised by the expert accounting witnesses in any further joint report pursuant to order 6, whichever is later. Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

The claims 1 These reasons for judgment explain why I have decided that the applicants' claims for patent infringement and misleading and deceptive conduct against the only remaining respondent should succeed and be subject to an award of damages calculated in a manner generally consistent with the approach of the accountant called by the applicants, but subject to a discount of 30% to take account of all relevant risks associated with the past hypothetical cash-flows which underlie the assessment of damages.

The essential facts 2 H Lundbeck A/S (referred to as Lundbeck A/S or Lundbeck DN) is the patentee of Australian patent 623144 (the 144 patent or the Lexapro patent). Lundbeck Australia Pty Ltd (referred to as Lundbeck AU), a subsidiary of Lundbeck A/S, claims to be the exclusive licensee of the 144 patent. Lundbeck AU sells Lexapro and Cipramil in Australia which are manufactured by and purchased from Lundbeck A/S. CNS Pharma Pty Ltd is a subsidiary of Lundbeck AU which sells a generic version of Lexapro in Australia known as Esipram which is also manufactured by and purchased from Lundbeck A/S. 3 Lundbeck A/S and Lundbeck AU, when there is no need to distinguish between them, are referred to in these reasons as Lundbeck. 4 The 144 patent is dated 13 June 1989 and thus its term of 20 years under s 67 of the Patents Act 1990 (Cth) meant that, if the term was not extended, the patent would expire on 13 June 2009. 5 Lundbeck A/S applied to extend the term of the 144 patent on 22 December 2003. The Commissioner of Patents granted the extension so that the 144 patent would not expire until 13 June 2014. Alphapharm Pty Ltd contended to the Commissioner that the extension was invalid and that the only extension which could have been granted was until 9 December 2012. The Commissioner agreed with Alphapharm's contention and scheduled a hearing to decide if the Commissioner should amend the extended term to 9 December 2012. 6 Alphapharm commenced a proceeding to revoke claims 1 to 6 of the 144 patent and to remove the extension of term until 13 June 2014 from the Register of Patents. Lundbeck commenced a proceeding to prevent the Commissioner from amending the Register by contending that a relevant regulation was invalid. Lundbeck also cross-claimed against Alphapharm for infringement of the 144 patent. 7 On 1 March 2006 Lindgren J dismissed Lundbeck's application contending invalidity of the regulation: H Lundbeck A/S v Commissioner of Patents [2006] FCA 163; (2006) 150 FCR 269. In so doing Lindgren J exposed the competing positions of the parties. Lundbeck relied on the registration of escitalopram oxalate on the Australian Register of Therapeutic Goods (ARTG) on 16 September 2003 as the relevant date for the extension. This is known as the Lexapro registration (Lexapro being the brand name of Lundbeck's escitalopram oxalate product). Alphapharm contended that the relevant date was that on which citalopram hydrobromide was registered on the ARTG, being 9 December 1997. This is known as the Cipramil registration (Cipramil being the brand name of Lundbeck's citalopram hydrobromide product). As Lindgren J explained at [13]: Subsection 71(2) of the Act provides that an application for an extension of term must be made during the term of the patent and within six months of the latest of the following dates: '(a) the date the patent was granted; (b) the date of commencement of the first inclusion in the [ARTG] of goods that contain, or consist of, any of the pharmaceutical substances referred to in subs 70(3); (c) the date of commencement of [s 71].' In this case, on either of the competing views the latest of those three dates is that referred to in para (b). If the date referred to in para (b) was, as Lundbeck contends, 16 September 2003, Lundbeck had until 16 March 2004 in which to apply for the extension, and therefore its application made on 22 December 2003 was within time. If, on the other hand, the date referred to in para (b) was, as Alphapharm contends, 9 December 1997, Lundbeck had only until 9 June 1998 in which to apply for an extension, and, therefore, its application on 22 December 2003 was made out of time. 8 On 19 May 2006 the Commissioner decided to amend the Register "to insert the correct extension of the term of the patent, that is 9 December 2012": Alphapharm Pty Ltd v H Lundbeck A/S [2006] APO 18; (2006) 69 IPR 629 at [35]. 9 In the meantime or thereafter, more proceedings were commenced. Arrow Pharmaceuticals Pty Ltd brought a claim for revocation of claims 1 to 6 of the 144 patent and rectification of the Register by removing the extension of term and recording that the 144 patent will expire on 13 June 2009 or, in the alternative, 9 December 2012. Sandoz Pty Ltd also brought a claim for revocation of claims 1 to 6 of the 144 patent and rectification of the Register by removing the extension of term. Sandoz's proceeding was settled and discontinued. Lundbeck A/S and Lundbeck AU entered into a deed of settlement with Sandoz in February 2007. 10 Lindgren J determined the Lundbeck, Alphapharm and Arrow proceedings on 24 April 2008: Alphapharm Pty Ltd v H Lundbeck A/S [2008] FCA 559; (2008) 76 IPR 618 (Lindgren J judgment). These proceedings are referred to in various documents and submissions as the Alphapharm proceedings. In short, Lindgren J dismissed Alphapharm's and Arrow's claims for revocation of all but claim 5 of the 144 patent and found that Alphapharm had infringed claims 1, 3, and 6 of the 144 patent. He also held that the extension of term was invalid as the relevant date under s 71(2) was the date of the Cipramil registration, with the consequence that reference to any extension of term of the 144 patent should be removed. 11 Lindgren J made orders on 19 June 2008. By order 6 Lindgren J ordered rectification of the Register to remove any reference to an extension of term of the 144 patent. By order 7, order 6 was stayed pending determination of an appeal to the Full Court on Lundbeck giving undertakings the terms of which are not now relevant. 12 On 11 June 2009 the Full Court (Emmett, Bennett and Middleton JJ) dismissed Lundbeck's appeal against order 6 concerning removal of any reference to an extension of term of the 144 patent. The majority (Bennett and Middleton JJ) also dismissed Alphapharm's and Arrow's claims for revocation of claims 1 to 6: see H Lundbeck A/S v Alphapharm Pty Ltd [2009] FCAFC 70; (2009) 177 FCR 151 (FFC#1). 13 The Full Court made orders on 12 June 2009 which stayed order 6 of Lindgren J's 19 June 2008 orders until determination of any application for special leave to appeal to the High Court on the basis of Lundbeck undertaking not, before determination of any such application, commencing or threatening proceedings against any person for infringement of the 144 patent in respect of conduct engaged in after 13 June 2009. 14 Accordingly, as at 12 June 2009: (1) The Full Court had held in FFC#1 that all purported extensions of term of the 144 patent, be it to 9 December 2012 or 13 June 2014, were invalid. (2) Order 6 had been made removing any reference to an extension of the term of the 144 patent from the Register. (3) Order 6 had been stayed on the basis of Lundbeck not commencing or threatening to commence infringement proceedings for conduct after 13 June 2009. (4) Without any extension of term of the 144 patent, that patent would expire at the end of its 20 year term on 13 June 2009. 15 Also on 12 June 2009 Lundbeck A/S applied for an extension of time in which to make an application to extend the term of the 144 patent to 9 December 2012 based on the Cipramil registration. For this purpose the time for making the extension of term application had to be extended to 12 June 2009, being the date on which Lundbeck A/S also applied for the extension of term of the 144 patent to 9 December 2012. 16 On 13 June 2009, pursuant to s 67 of the Patents Act, the 20 year term of the 144 patent ended. 17 From Monday, 15 June 2009 onwards, Apotex Pty Ltd, Sigma Pharmaceuticals (Australia) Pty Ltd and Sandoz supplied generic escitalopram oxalate products in Australia. 18 On 11 December 2009 the High Court refused the applications for special leave to appeal. 19 On 9 February 2010 the Register was rectified to remove reference to the extension of term of the 144 patent. 20 Alphapharm, Apotex and Sandoz opposed the extension of time application before the Commissioner, as did Sigma. Sigma was subsequently acquired by the Aspen group of which Aspen Pharma Pty Ltd is part. 21 On 1 June 2011 the Commissioner decided to extend the time for the making of an application to extend the term of the 144 patent based on the Cipramil registration until 9 December 2012: Alphapharm Pty Ltd v H Lundbeck A/S [2011] APO 36; (2011) 92 IPR 628. 22 Alphapharm, Apotex, Sandoz and Aspen appealed against the Commissioner's decision to the Administrative Appeals Tribunal (the AAT). On 4 December 2012, the AAT affirmed the Commissioner's decision granting the extension of time: Aspen Pharma Pty Ltd v Commissioner of Patents [2012] AATA 851; (2012) 132 ALD 648. 23 On 18 November 2013 the Full Court dismissed the appeals against the AAT's decision: Aspen Pharma Pty Ltd v H Lundbeck A/S [2013] FCAFC 129; (2013) 216 FCR 508 (FFC#2). 24 In the meantime Alphapharm, Apotex, Sandoz and Aspen had filed oppositions to the extension of term of the 144 patent and had applied under s 223(9) of the Patents Act for licences to exploit the 144 patent during its extended term, should its term be extended. 25 On 25 June 2014 the Commissioner granted the extension of term of the 144 patent to 9 December 2012: Alphapharm Pty Ltd v H Lundbeck A/S [2014] APO 41; (2014) 109 IPR 323. 26 On 26 June 2014 Lundbeck commenced infringement proceedings against Alphapharm, Apotex, Sandoz and Aspen. On 30 May 2016 CNS Pharma also commenced proceedings against those entities seeking declarations that, by the sale, offering for sale and supply of escitalopram products during the term of the 144 patent, they engaged in misleading and deceptive conduct. Accordingly, CNS Pharma sought damages for the loss it claims to have suffered during the patent term as a result of the alleged acts of infringement by the generic parties. 27 Having granted special leave to appeal, on 5 November 2014 the High Court dismissed Alphapharm's appeal against the Full Court's orders dismissing the appeal to it by a 3:2 majority: Alphapharm Pty Ltd v H Lundbeck A/S [2014] HCA 42; (2014) 254 CLR 247. 28 Alphapharm, Apotex, Sandoz and Aspen appealed against the Commissioner's decision to extend the term of the 144 patent and, on 6 November 2014, Rares J dismissed this appeal: Alphapharm Pty Ltd v H Lundbeck A/S [2014] FCA 1185; (2014) 110 IPR 59 (the Rares J judgment). 29 Alphapharm, Apotex, Sandoz and Aspen appealed against Rares J's orders dismissing the appeal and on 22 September 2015 the Full Court dismissed that appeal: Alphapharm Pty Ltd v H Lundbeck A/S [2015] FCAFC 138; (2015) 234 FCR 306 (FFC#3). 30 Alphapharm, Apotex, Sandoz and Aspen applied for, but on 11 March 2016 were refused, special leave to appeal to the High Court against the orders in FFC#3. 31 On 5 August 2016 Lundbeck A/S applied for an order that the Commissioner had no power to determine the applications for licences under s 223(9) of the Patents Act. 32 On 21 October 2016 I held that Alphapharm and Aspen were not estopped or precluded by the doctrine of abuse of process from raising the status of Lundbeck AU as an exclusive licensee or not of the 144 patent: H Lundbeck A/S v Alphapharm Pty Ltd [2016] FCA 1232. 33 On 3 February 2017 Beach J held that the Commissioner did have the power to determine the licence applications and made declarations reflecting this conclusion: H Lundbeck A/S v Commissioner of Patents [2017] FCA 56; (2017) 249 FCR 41. 34 Before the first tranche of the hearing of the current matters commenced, the claims against and by Alphapharm were discontinued. The claim against Aspen was also subsequently discontinued after the first tranche of the hearing. During the second tranche of the hearing, the claims between Lundbeck and Apotex settled. As a result, the sole remaining respondent is Sandoz. 35 The fact that Sandoz alone remains as a respondent creates some practical difficulty. To explain, Sandoz did not challenge the validity of the 144 patent, but it did defend Lundbeck's infringement proceedings on various grounds including the proper construction of the 144 patent. In so doing Sandoz relied on Apotex's expert evidence and submissions. Apotex's submissions about construction were part of its case that the 144 patent is invalid or, if valid, should be construed in a manner which meant that Apotex's products (and thus also Sandoz's products) did not infringe. As noted, however, the proceedings against Apotex settled, but only after the hearing in relation to the invalidity and the proper construction of the 144 patent. 36 I sought clarification of Sandoz's position. Sandoz said that it pressed its non-infringement argument based on the proper construction of the 144 patent as articulated by Apotex, but did not contend the patent was invalid as a result of any particular construction. 37 The difficulty is that given that the hearing took place in the way that it did it is not possible to identify the construction argument other than in the context of Apotex's invalidity argument. As a result, in these reasons for judgment I refer to and deal with Apotex's invalidity argument, but I do so only for the purpose of explaining my conclusions about Sandoz's construction and associated non-infringement argument. Sandoz's construction argument relied on one aspect of Apotex's case, which may be described as the "separated, isolated or pure" (+)-enantiomer argument. Sandoz did not adopt Apotex's so-called "molecule" construction argument. However, in order to understand the competing constructions, and thus Sandoz's position as the sole remaining respondent, it is necessary to explain all aspects of Apotex's arguments. 38 Sandoz also adopted other submissions of various generic parties about certain issues. For ease of reference I have attributed all of those submissions in these reasons to Sandoz but, in reality, many of the submissions were made by one or other generic party before the cases against them were settled and which Sandoz adopted.

Apotex's challenge to validity of the 144 patent

3.1 The grounds of the challenge 39 During the first tranche of the hearing, Apotex contended that claims 1 and 3 of the 144 patent were invalid on three grounds. 40 The first ground was that the invention claimed in claims 1 and 3 was not a patentable invention as the invention was not novel as required by s 18(1)(b)(i) of the Patents Act by reason of the publication of Australian Patent No. 509445 (the 445 or Citalopram patent) on or about 13 July 1978. 41 The second ground was that if claim 1 of the 144 patent, on its proper construction, is limited to the "separated or isolated or pure (+)-enantiomer of 1-(3-dimethylaminopropyl)-1-(4'-fluorophenyl)-1,3-dihydroisobenzofuran-5-carbonitrile with nothing else present", then the invention claimed in claims 1 and 3 is not fairly based on matter described in the specification as required by s 40(3) of the Patents Act. 42 The third ground was that if claim 1 of the 144 patent, on its proper construction, is not so limited then the term "(+) enantiomer of 1-(3-dimethylaminopropyl)-1-(4'-fluorophenyl)-1,3-dihydroisobenzofuran-5-carbonitrile" in claim 1 is not clear and claims 1 and 3 therefore do not comply with s 40(3) of the Patents Act. 43 Although Apotex's cross-claim referred to the 1990 Patents Act, as it explained in its written submissions, validity, but not infringement, is to be decided under the provisions of the Patents Act 1952 (Cth) (the 1952 Act). This was not in dispute so it suffices to adopt Apotex's submission about this issue as follows: Lindgren J [in the Lindgren J judgment] explained the continued relevance of the Patents Act 1952 (Cth) (1952 Act) at [54]-[55]. This was clarified by Middleton J in Eli Lilly and Company Limited v Apotex Pty Ltd [[2013] FCA 214] (2013)100 IPR 451 at [7]-[18], referring in particular to reg 23.26 of the Patents Regulations 1991, and to ss 233 and 234 of the Patents Act 1990 (Cth) (Act). For the reasons noted by Middleton J at [16], by a combination of s 228(7) and reg 23.26(2), when a patent has been granted under the Act on an application made under the 1952 Act, the 1952 Act applies to validity but not to infringement. 44 To the extent relevant to the present matter, the requirement that an invention be novel, and that claims be clear and fairly based on the matter described in the specification under the 1952 Act involve the same considerations as under the 1990 Act. 45 It is unnecessary to record the uncontroversial scientific facts about stereochemistry, enantiomers, citalopram and escitalopram. There is no material dispute about these matters. 46 Nor is it necessary to summarise the 144 and 445 patents. This exercise has been done repeatedly in the previous decisions, which are an essential part of the background to the current dispute.

3.2 Construction of claim 1 47 The Lindgren J judgment identifies the background to the Alphapharm proceedings in uncontroversial terms at [1] to [8]. His Honour said: 1 These four proceedings, which were heard together, relate to Australian patent No 623144 (the Patent or the Australian Escitalopram Patent) held by H Lundbeck A/S (Lundbeck), a Danish pharmaceutical company. 2 Lundbeck applied for the Patent on 13 June 1989. The application was a Convention application and was said in the application to be based on application No 8814057 for a patent made in the United Kingdom on 14 June 1988 (the UK Escitalopram Patent - see [79] below). The title of the invention in the Patent is: "(+)-Enantiomer of citalopram and process for the preparation thereof". 3 Citalopram is a molecule patented by Lundbeck which is used for the treatment of depression. Citalopram is a chiral molecule, a racemic mixture (or racemate) comprising in equal measure two enantiomers. Enantiomers are non-superimposable mirror images of each other. They are designated "(+)" or "(-)" based on a particular physical property referred to below, and "R" or "S" based on their three-dimensional structure. The correlation between the (+) or (-) and the R or S designations can only be determined, however, through experimentation. 4 In the case of citalopram, experimentation subsequent to the priority date has shown that the (+)-enantiomer is in fact the S-enantiomer. It is commonly referred to as "S-citalopram", and has the International Nonproprietary Name "escitalopram". 5 The Patent discloses processes for obtaining escitalopram, and data showing that (+)-citalopram is therapeutically more active than citalopram itself, and more than 100 fold more active than (-)-citalopram. 6 The 20 year term of the Patent was due to expire on 13 June 2009. The term has, however, been extended as discussed at [28] ff below. 7 Citalopram is an invention claimed in Australian patent No 509,445 (the Australian Citalopram Patent) dated 5 January 1977, the term of which was sixteen years commencing on that date. The Australian Citalopram Patent was, in turn, said in the application to be based on the application No 1486/76 for a patent filed in Great Britain on 14 January 1976 (the UK Citalopram Patent). The title of the invention in the Australian Citalopram Patent is "phthalanes", a class of compounds that includes citalopram. 8 In various ways, the proceedings before the Court raise issues concerning the relationship between citalopram and escitalopram. Broadly, the issues can be separated into those that relate to patentability (raised in the proceedings brought by Alphapharm Pty Ltd (Alphapharm) and Arrow Pharmaceuticals Pty Ltd (Arrow) for revocation of the Patent), infringement (a cross-claim in Alphapharm's revocation proceeding) and regulatory aspects (all four proceedings). 48 Lindgren J identified the claims of the 144 patent as follows: 9 The Patent comprises six claims. … 10 The six claims are as follows: 1. (+)-1-(3-dimethylaminopropyl)-1-(4'-fluorophenyl)-1,3-dihydroisobenzofuran-5-carbonitrile and non-toxic acid addition salts thereof. 2. The pamoic acid salt of (+)-1-(3-dimethylaminopropyl)-1-(4'-fluorophenyl)-1,3-dihydroisobenzofuran-5-carbonitrile. 3. A pharmaceutical composition in unit dosage form comprising as an active ingredient, a compound as defined in claim 1, together with a pharmaceutically accepta[ble] carrier or excipient. 4. A pharmaceutical composition in unit dosage form comprising, as an active ingredient, the compound of claim 2, together with a pharmaceutically acceptable carrier or excipient. 5. A pharmaceutical composition in unit dosage form, according to claim 3 or 4, wherein the active ingredient is present in an amount from 0.1 to 100 milligram per unit dose, together with a pharmaceutically acceptable carrier or excipient. 6. A method for the preparation of the compound of claim 1, which comprises: (a) reacting a compound of the formula with an enantiomerically pure acid derivative as an acid chloride, anhydride or labile ester, subsequently separating the resolved intermediate enantiomer for (+)-citalopram having the formula wherein R is a labile ester group; by HPLC or fractional crystallization, and then treating said intermediate enantiomer with strong base: or (b) reacting a compound of formula II with the enantiomer of an optically active acid affording the pure enantiomer salt of the compound of formula II for (+)-citalopram, and subsequently performing ringclosure via a labile ester by reacting the pure enantiomer of formula II as a base with an activated acid with simultaneous addition of a base and, if desired, transferring the (+)-citalopram obtained to a pharmaceutically acceptable salt thereof. 49 Lindgren J construed claim 1 of the 144 patent to mean (+)-citalopram. He stated his conclusion as follows: 116 The symbol (+)- in claim 1 (and in claim 2) indicates the particular enantiomer that is distinguished by the feature of rotating plane-polarised light to the right (under standard conditions - see Section G below). The property of rotating plane-polarised light to the right (clockwise) or to the left (anticlockwise) which distinguishes one enantiomer from the other and each from the racemate, is able to be detected only when that enantiomer is present other than as a part of the racemate. This is because when in a racemic mixture, the polarisation of light by each enantiomer is "cancelled out" by the other enantiomer, and there is thus no net polarisation of light. The skilled addressee in 1988 would have understood claim 1 to refer specifically to (+)-citalopram. 117 In the absence of a context permitting otherwise, the skilled addressee would not have understood the claim to the invention of (+)-citalopram to refer to that compound merely as part of the unresolved racemate. Rather, the skilled addressee would have understood the claim to the invention of (+)-citalopram to refer to that compound as something that had an existence independent of that of the racemate. 50 His Honour's reasoning included the following: 118 The only context that I take into account in arriving at this conclusion is that it was part of common general knowledge: • that racemates contained in equal parts (+) and (-) enantiomers; • that the enantiomers of a racemate were potentially separable, and it was a possibility that an enantiomer might have a stable existence as a compound distinct from being part of the racemate; • that some racemates had in fact been resolved into their separate enantiomers; and • that racemates were commonly represented by the symbol (±) to indicate the presence of both enantiomers. 119 It is not that the (+)-citalopram compound does not exist when it is part of the racemate. The skilled addressee would understand that it exists whether a part of the racemate or apart from it, and that the purpose of the (+) symbol is only to distinguish it from the (-)-enantiomer and from its being merely part of the racemate or of some other mixture. The (+) or (-) symbol, devoid of any context suggesting otherwise, implies distinctness from the racemate. … 123 So, the unqualified reference to the (+)-enantiomer of citalopram in claim 1 refers to something different from that enantiomer as an indistinguishable part of the unresolved racemate. 51 Lindgren J noted the expert evidence relevant to construction at [128]-[140], which founded his Honour's conclusion at [138] that "chemists at the priority date and now would understand the bare formula of claim 1 to require a compound that was at least 95% pure (+)-citalopram". Otherwise, in reaching his conclusions as to the proper construction of claim 1, Lindgren J had regard to the body of the specification at [141]-[142]. After rejecting various submissions as to the relevance of other matters, Lindgren J said at [153]: In summary, the reference in claim 1 to (+)-1-(3-dimethylaminopropyl)-1-(4'-fluorophenyl)-1,3-dihydroisobenzofuran-5-carbonitrile is a specific reference to the enantiomer of citalopram that rotates plane-polarised light to the right, and, there being no context suggesting a different possibility, is not apt to refer to the (+)-enantiomer merely as present in the racemate. 52 In FFC#1 Bennett and Middleton JJ were in the majority and Emmett J was in dissent. 53 Apotex argued that the decision of the majority about construction of claim 1, and that of Lindgren J, depended on expert evidence. Thus Apotex pointed to Bennett J's observation in FFC#1 at [150] that the "[t]he primary judge was entitled to accept the evidence that the skilled addressee would read claim 1 as referring to the isolated (+)-enantiomer. In the context of the whole of the specification, this was in contrast to the (+)-enantiomer as present in the racemate or in a mixture with the (-)-enantiomer" and to what Middleton J said at [252]: Claim 1 simply identifies the substance (+)-enantiomer. The construction I prefer does not involve reading into claim 1 a limitation of "independently existing" or "existing independently" of the racemate any more than it expands the context in which the (+)-enantiomer is to be found. The mere reference to (+)-enantiomer in claim 1 without more indicates that it is the separated or the isolated enantiomer that is claimed. In other words, it is implicit in the way claim 1 is worded, where the Patent is entitled "(+)-Enantiomer of Citalopram and process for the preparation thereof", that the claim is only to the isolated or separated (+)-enantiomer. 54 I am not persuaded that the Lindgren J judgment or FFC#1, insofar as the issue of construction is concerned, depended on expert evidence. As Lindgren J's reasons disclose, he relied on certain contextual matters at [118], but those matters were not in dispute on the evidence. He also had regard to the body of the specification. But his conclusion did not depend on resolving any contested issue of expert evidence. Nor should anything in the reasoning of Bennett or Middleton JJ in FFC#1 be understood as suggesting to the contrary. 55 To the extent [131]-[138] of Bennett J's reasons were relied upon by Apotex, it is apparent that her Honour was recording common ground other than with respect to the 95% purity issue to which Lindgren J referred at [138]. 56 Given this, I am not persuaded by Apotex's submissions that the construction of claim 1 reached in FFC#1 depended on expert evidence, with the consequence that I am not bound by that construction. To the contrary, I consider Lundbeck to be correct that claim 1 has been conclusively construed in FFC#1 as Lindgren J proposed in [153] of the Lindgren J judgment. In reaching this conclusion Bennett J, having confirmed that Lindgren J was correct at [150], said this at [151]: That is, claim 1 is to the separated or isolated or pure (+)-enantiomer. 57 These words "separated or isolated or pure" have taken on their own life in this matter but it is best to understand now that Bennett J was doing nothing more than affirming that Lindgren J's conclusion at [153] was correct. 58 In FFC#3, to which Apotex and Sandoz were parties, the Full Court (Bennett, Nicholas and Yates JJ) had to revisit the issue of construction. FFC#3 involved an appeal against the Rares J judgment in which Rares J had dismissed the generic parties' appeal against the Commissioner's decision extending the term of the 144 patent until 9 December 2012. Rares J recorded the generic parties' case on appeal at [6] of the Rares J judgment: (1) the pharmaceutical substance per se disclosed in claim 1 of the patent was the pure, isolated or separated form of the molecule, being the (+)-enantiomer, for the purposes of s 70(2) of the Act, and that that pure form of the molecule was not included in the goods registered as "CIPRAMIL Citalopram hydrobromide 20mg tablet blister pack" on the ARTG with a start date of 9 December 1997 for the purposes of s 70(3). The Cipramil tablets were produced as an exploitation of the properties of the racemate (the per se issue); (2) the requirement of s 70(4) of the Act, that the term of the patent had not previously been extended, could not be satisfied because the term of the patent in suit had been previously extended to 13 June 2014 by the delegate and that extension had been recorded in the Register of Patents on 17 June 2004, although, subsequently, on 19 June 2008, Lindgren J had ordered under s 192 of the Act that the Register be rectified by removing the extension of time, and the Full Court upheld that decision (the prior extension issue); (3) the delegate erred because he exceeded his function of deciding the opposition under s 75(2) by also proceeding to grant the challenged extension of time under s 76(1)(b) of the Act, before this appeal had been determined (the erroneous grant issue). 59 Section 70 of the Patents Act, as referred to in these paragraphs provides that: (1) The patentee of a standard patent may apply to the Commissioner for an extension of the term of the patent if the requirements set out in subsections (2), (3) and (4) are satisfied. (2) Either or both of the following conditions must be satisfied: (a) one or more pharmaceutical substances per se must in substance be disclosed in the complete specification of the patent and in substance fall within the scope of the claim or claims of that specification; (b) one or more pharmaceutical substances when produced by a process that involves the use of recombinant DNA technology, must in substance be disclosed in the complete specification of the patent and in substance fall within the scope of the claim or claims of that specification. (3) Both of the following conditions must be satisfied in relation to at least one of those pharmaceutical substances: (a) goods containing, or consisting of, the substance must be included in the Australian Register of Therapeutic Goods; (b) the period beginning on the date of the patent and ending on the first regulatory approval date for the substance must be at least 5 years. (4) The term of the patent must not have been previously extended under this Part. … 60 In FFC#3 the Full Court had to deal with the Lindgren J judgment and FFC#1 (referred to in FFC#3 as the Decision). After referring to claim 1 of the 144 patent, the Full Court in FFC#3 said at [15]: In essence, in the Decision, the majority concluded that the claim was to the separated enantiomer and that the claim was not anticipated by the Cipramil Patent which described the racemate, as the racemate did not describe or disclose 'the pure or isolated (+)-enantiomer', nor was there anything to tell the skilled addressee to resolve the racemate into the two enantiomers, or how to do so. The majority also concluded that the racemate "contained" the (+)-enantiomer for the purposes of s 70(3)(a) of the Act. 61 The Full Court also referred in [18] to "some relevant aspects of the background chemistry which are not in dispute", in these terms: • (+)-citalopram is the (+)-enantiomer of citalopram; • (+)-citalopram is also known as (S)-citalopram or escitalopram; • Citalopram is a racemic mixture that contains equal amounts of the (+) and (-) enantiomers of citalopram; • Citalopram and escitalopram are different chemical entities with different physical properties and different pharmacological activities. 62 The undisputed nature of this evidence is confirmed by the following statement of the Full Court at [19]: The background chemistry has been described in some detail in the Decision and does not need to be repeated. 63 This accords with my conclusion above that, insofar as expert evidence was relevant to claim 1, it was evidence that was not in dispute in FFC#1. 64 The Full Court also rejected the arguments of the generic parties (repeated here by Apotex) about the reasoning in FFC#1. To understand this aspect of the matter, it is necessary to record the submissions which were rejected in FFC#3. The Full Court recorded the following: 47 Alphapharm accepts that the (+)-enantiomer is a pharmaceutical substance. The essence of its argument is that in the Decision, the subject matter of the claim was construed as the separated, isolated or pure enantiomer, which introduces a further criterion or limitation such that the claim is no longer a claim to the enantiomer per se. This is because the requirement that it be the substance per se means that there must be no further criterion (News Limited v South Sydney District Rugby League Football Club Limited (2003) 215 CLR 563 at [56]). Of course, it says, Cipramil cannot "contain" the separated molecule, the molecule free of the racemate, so it follows that what is contained in Cipramil is not the subject matter of the claim. Putting Alphapharm's submissions another way, for the pharmaceutical substance per se the subject of the claim to comply with the requirements of s 70(2)(a) of the Act, it would be the isolated enantiomer and there are no goods included in the ARTG containing the pharmaceutical substance as required by s 70(3)(a) because the racemate contains the (+)-enantiomer and the (-)-enantiomer together. 48 Accordingly, Alphapharm submits, Lundbeck is not entitled to an extension of term of the Patent. 49 Alphapharm does not challenge the construction of the claim as adopted by the majority in the Decision. It says that the construction of the claim means that s 70(2)(a) is not satisfied. It points to the conclusion in the Decision that the racemate did not anticipate the separated enantiomer because it did not describe or disclose the separated or isolated or pure enantiomer and says that this means that it could not contain that enantiomer. 50 The issue, as distilled in this appeal, is whether the Decision demonstrates that the claim was held to include the criterion that the (+)-enantiomer was isolated or pure, either as an additional integer or by way of limitation. If so, Alphapharm contends, the claim is not to the molecule or to the compound. If it were, it says, the Lundbeck Full Court would have found, as did Emmett J who characterised the claim as a claim to the molecule, that it was anticipated by the Cipramil Patent which claimed the racemate. 65 The Full Court rejected these submissions including in the following paragraphs: 72 At [151], Bennett J said: '[t]hat is, [the claim] is to the separated or isolated or pure (+)-enantiomer'. It is apparent that, in context, her Honour was not considering levels of purity, as is made clear from the Decision at [157] to [160]. At [157] Bennett J expressly rejected the suggestion that the claim imported a limitation of purity. In that context her Honour drew a distinction between a limitation of purity and an 'isolated, separated enantiomer, not in a mixture or in a racemate'. 73 It is quite clear that those words "isolated" and "pure" were not intended to add a further qualification or limitation to the claim. Rather, they imported the concept of "separate", as is evident from [149] where her Honour said: The invention is also said to be concerned with a method to resolve the racemate into the individual isomers. While the specification refers to the enantiomers, in the plural, it is in the context of their separate identity. 74 Her Honour then referred to the evidence that the skilled addressee would read the claim as referring to the isolated (+)-enantiomer, in contrast to the (+)-enantiomer as present in the racemate or in a mixture with the (-)-enantiomer. Clearly, in context, the word "isolated" referred to a resolution of the racemate into the separate enantiomers… 75 Turning to the alleged anticipation by the racemate disclosed and claimed in the Cipramil Patent, Bennett J said (at [194]) that while the skilled addressee knew that the racemate could be resolved into enantiomers, there was nothing to tell him or her to do so. Her Honour also noted that the Cipramil Patent was silent as to the means of obtaining the enantiomers. That is, Bennett J drew a distinction between the enantiomers unseparated or unresolved in the racemate and separated to yield the (+)-enantiomer of the claim. Justice Bennett took the view that the question that must be asked is whether Cipramil contained the same pharmaceutical substance per se disclosed and claimed in the Patent, namely the (+)-enantiomer. 76 Justice Bennett agreed with Lindgren J that the racemate was a good that contained the pharmaceutical substance (+)-citalopram. This reasoning constituted acceptance of (+)-citalopram as a pharmaceutical substance per se. … 78 The construction of the claim, as understood by the skilled reader, does not import a limitation of purity. This is also made clear in the reasons of Middleton J. His Honour agreed with the reasons of Bennett J and dealt specifically with the submission that Lindgren J had impermissibly added integers that are not found in the claim. His Honour said at [252] that the claim 'simply identifies the substance (+)-enantiomer'. 79 This construction did not involve reading into the claim a limitation of "independently existing" or "existing independently" of the racemate any more than it expands the context in which the (+)-enantiomer is to be found. Any reference to the (+)-enantiomer in the claim indicates that it is the separated or isolated enantiomer that is claimed. In other words, it is implicit in the way that the claim is worded, where the Patent is entitled '(+)-enantiomer of citalopram and processes for the preparation thereof', that the claim is only to the isolated or separated (+)-enantiomer. 80 Accordingly, on the construction of the claim determined in the Decision, there is no further limitation imported into that claim. That was the clear conclusion of Bennett and Middleton JJ in the Decision. It is also consistent with their Honour's reasoning and conclusion as to s 70 of the Act and as to whether citalopram, the racemate, contained the pharmaceutical substance per se, being the (+)-enantiomer or escitalopram disclosed and claimed in the Patent. 66 Turning to the submissions in FFC#3 the Full Court said this: 82 Alphapharm submits that, because of the evidence accepted by Lindgren J that a skilled person looking at the structure of citalopram would have recognised the presence of two enantiomers and would have been able to depict them, a claim to the enantiomer per se would necessarily have been invalid. Alphapharm then submits that in order to find that the (+)-enantiomer citalopram was novel, Lindgren J and Bennett and Middleton JJ construed the claim as being not to the enantiomer itself but rather to the enantiomer further defined or limited by the requirement that it be separated or isolated or pure. This ignores the wording of the claim, which does not import, in terms, the limitation that it be "isolated" (cf D'Arcy v Myriad Genetics Inc (2014) 313 ALR 627 at [1]). 83 It also ignores the reasoning of each of Lindgren J, Bennett and Middleton JJ in concluding that the racemate did not anticipate the (+)-enantiomer. This was not because the claim should be read as being to the isolated enantiomer but, as explained in the Decision at [193]-[195], the disclosure of the racemate was not a disclosure to the skilled addressee of the (+)-enantiomer. As explained at [193], the skilled addressee would have understood that (±)-citalopram (the racemate) consisted of the (+)-enantiomer and the (-)-enantiomer and would have been able to identify the formulae for the enantiomers but would not have known in the absence of experimentation which was the (+)-enantiomer and which was the (-)-enantiomer. That is, there was no prior disclosure of the enantiomer independently existing. As Bennett J explained at [194], the skilled addressee knew that the racemate could be resolved into the enantiomers but there was nothing to tell him or her to do so. It could not even be said that there were clear and unmistakable directions to obtaining the enantiomers. Accordingly, it does not follow that a claim to the enantiomer per se would necessarily have been invalid. … 88 Alphapharm also contends that Lundbeck had submitted in the Lundbeck Full Court that the requirement of 'separated or isolated or pure' was part of the text of the claim perceived by reading 'the (+) symbol'. It relies upon the fact that this submission was accepted by Lindgren J at [119] and [123] and on the emphasis on the (+) symbol in Bennett J's reasons. However, again, Alphapharm ignores Bennett J's reasoning generally and specifically in rejecting a limitation as to purity (at [157]). Similarly, Middleton J stated that he was not reading into the claim a limitation of "independently existing" (at [252]). However, Alphapharm submits that this is because Lindgren J and Bennett and Middleton JJ considered that the integer or limitation did not need to be read into the claim as it was already part of the claim. Alphapharm also submits that separation, isolation and purification each refers to a process so that the claim is not to the enantiomer in itself, which it repeats would not have been novel, but only to the enantiomer when separated or isolated or purified. … 91 In seeking to describe the difference between the racemate containing the two enantiomers and the claim to one of the enantiomers, Bennett and Middleton JJ used expressions such as "pure", "isolated" and "separated", the same language used by Lindgren J. Another way of describing the subject matter of the claim was 'specific reference to the (+)-enantiomer, distinct from the (-)-enantiomer and not as present in the racemate mixture' (the Decision at [131]). In seeking to differentiate between the racemate and the (+)-enantiomer, the word "itself" could just as easily have been used and would have carried the same meaning in context. 92 Justices Bennett and Middleton did not read into the claim an additional integer, nor did they add something to the words of the claim. Their reasoning was that the claim is to the (+)-enantiomer and nothing else. The term "pharmaceutical substance per se" simply means the pharmaceutical substance "in itself". In Boehringer, Heerey J observed that per se meant 'by or in itself' 'intrinsically' or 'essentially' (at [7]). The Full Court approved that approach on appeal (at [37]). 93 Semantics aside, it is clear that the claim describes a pharmaceutical substance per se. The substance was, as explained by Lindgren J at [536], a new chemical entity. The racemate and the (+)-enantiomer had different physico-chemical interactions manifested in different pharmacodynamics and pharmacokinetics (the Decision at [234]). The claim is to (+)-citalopram, irrespective of how it is produced. The isolated (+)-enantiomer plainly qualifies as a pharmaceutical substance per se and the primary Judge was correct in concluding that it satisfies s 70(2)(a) of the Act. 94 As pointed out by Lundbeck, Alphapharm's submission that Bennett and Middleton JJ construed the claim to include an additional integer or a limitation, such that the claim was no longer to the (+)-enantiomer as a pharmaceutical substance per se, is wholly inconsistent with their Honours' conclusion that the subject matter of the claim satisfied s 70(2)(a) of the Act, such that Lundbeck was entitled to an extension of term based upon Cipramil but not on Lexapro. … 96 It is clear from the Decision that the expressions such as "separated or isolated or pure" did not import any concept of degree of purity. Quite clearly, Bennett J used the expression as a hendiadys, to explain that the (+)-enantiomer alone was claimed - without the (-)-enantiomer or the racemate. That is, it was the enantiomer itself. Put another way, it was the enantiomer per se. 67 In short, I agree with Lundbeck that Apotex and Sandoz, as parties to FFC#3, are subject to an issue estoppel with respect to the proper construction of claim 1 of the 144 patent. Apotex (and thus Sandoz) is not at liberty to argue, as they did, that claim 1 should be construed as Emmett J, in dissent proposed, in FFC#1. The construction of claim 1 in accordance with FFC#1 and FFC#3 was an essential or indispensable step leading to the dismissal of the appeal in FFC#3. The construction was not based on disputed expert evidence but uncontentious common ground between the parties as to the background science and the terms of the claim construed in the context of the specification as a whole. 68 Even if these conclusions were incorrect, Apotex's argument that the evidence is different in the present case is unpersuasive. In short, Apotex relies on the evidence of the experts whom Lundbeck called, Professor Stephen Davies and Dr Alan Robertson, about infringement of the 144 patent to support its case. As Apotex argued, Professor Davies and Dr Robertson, in their evidence as to infringement, looked at the relevant information to identify the mere presence of the (+)-enantiomer. According to Apotex, this is "powerful evidence" that the construction of claim 1 in the Lindgren J judgment, FFC#1 and FFC##3 is wrong and that, rather, claim 1 claims the molecule, the (+)-enantiomer whether or not that molecule is in a mixture, racemic or otherwise, and irrespective of the degree of purity of the mixture. 69 I do not accept these submissions. Their evidence, taken as a whole, is inconsistent with these propositions. It was not put to Professor Davies or Dr Robertson during the course of these proceedings that the evidence they gave about infringement was inconsistent with the evidence they gave to the effect that they understood claim 1 to mean the (+)-enantiomer separate from the racemic mixture. The only difference between their evidence, and the approach in FFC#3 was that, in common with Lindgren J, they remained of the view that claim 1 would be read by the person skilled in the art as permitting chiral impurity provided that 95% purity was achieved. Bennett J (and thus Middleton J) rejected this in FFC#1 at [154]-[160]. In so doing, however, their Honours did not suggest that 100% chiral purity was required. 70 Apotex's arguments to the contrary are not persuasive. As Lundbeck submitted, it was common ground that it is not possible to separate a racemic mixture into 100% pure (+)-enantiomer and (−)-enantiomer. This fact did not mean that Professor Davies and Dr Robertson accepted that claim 1 claimed the molecule, the (+)-enantiomer whether or not it exists in a mixture. Their evidence was to the opposite effect, that claim 1 refers to the separate enantiomer which will not be 100% pure because 100% separation cannot be achieved. When dealing with the issue of infringement, to the contrary of Apotex's submission, their evidence is not to be understood as focusing on the mere presence of the (+)-enantiomer in the alleged infringing products. Their focus was the presence of the separated (+)-enantiomer which, as the evidence disclosed, was the active ingredient in the respondent's allegedly infringing products to a very high level of purity. Indeed, the evidence was that it would be near impossible to achieve chiral impurity of the kind posited by Apotex (6% (−)-enantiomer) without simply adding back the requisite amount of the racemic mixture. As Dr Robertson said, to separate the (+)-enantiomer to the high level of purity possible (over 99%) then to add back in the racemate to achieve a lower level of purity would be perverse. Yet Apotex's approach to construction depends on this kind of reading of the 144 patent, as well as FFC#1 and FFC#3 (despite, it must be said, FFC#3 containing a clear and cogent explanation of how FFC#1 cannot be understood as was proposed in that case and as Apotex proposes again in the present case). 71 Apotex's contrary submissions relied on propositions which, in the circumstances had to be, but were not put to Professor Davies or Dr Robertson or otherwise focus on aspects of their evidence taken out of context. For this reason the fact that Dr Robertson did not give evidence before Lindgren J is immaterial. His evidence, understood fairly and in context, did not support Apotex's molecule construction. The submission unfairly confuses Dr Robertson's evidence identifying the (+)-enantiomer as the "particular molecule" with Apotex's case that claim 1 means the (+)-enantiomer whether or not in a mixture. It is plain that Dr Robertson did not accept any such suggestion and that he understood claim 1 to mean the separated (+)-enantiomer (albeit that it could never be 100% pure) and not the (+)-enantiomer within a racemic mixture. Nothing Dr Robertson said in the joint report supports the contrary proposition put by Apotex. 72 Nor was there any material difference between the evidence given by Professor Davies before Lindgren J and in the present case. He too remained of the view that claim 1 had nothing to do with the (+)-enantiomer in the racemic mixture. He cannot be fairly criticised for taking into account the Full Court's rejection of his view that claim 1 would be understood as involving a purity of at least 95% in FFC#1. It became apparent that Professor Davies' view remained that a skilled addressee would understand claim 1 as involving the separated (+)-enantiomer to at least 95% purity. 73 It is unnecessary to address the evidence of Dr Richard Oppenheim on these matters in detail. Dr Oppenheim agreed that the (+)-enantiomer could not be separated from the racemate to achieve 100% pure (+)-enantiomer. He, wrongly, believed that FFC#1 required claim 1 to be understood as requiring 100% purity, relying on Bennett J's reference to "separated or isolated or pure". This belief is not supported by a reasonable reading of FFC#1 and, on no view, can survive the reasoning in FFC#3 which rejected this very proposition. 74 Leaving aside all these matters, Apotex did not confront the fact that the same construction issue has been the subject of extensive consideration in the Lindgren J judgment, FFC#1 and FFC #3. The argument that claim 1 means the molecule, the (+)-enantiomer whether or not it exists within the racemic mixture has been repeatedly rejected. The further argument (which Sandoz adopted) that, assuming this to be so, claim 1 means the (+)-enantiomer in a form which is 100% pure was also comprehensively rejected in FFC#3 and, in any event, was not open on a fair reading of FFC#1. Given this, why a different approach should be taken in the present matter is not apparent. The submission that the previous decisions relied on expert evidence which is different in the present case, as noted, is without real substance. Ultimately, Apotex (and thus Sandoz) gave no cogent reason to support its case on construction but merely ran the same arguments which have bene repeatedly rejected in the previous decisions. 75 Once this is recognised, Apotex's other arguments, based on the specification as filed and as amended, provide no support to its case. The point Apotex attempted to make, that where the specification intends to refer to purity it does so expressly, involves a premise about claim 1 which has been repeatedly rejected, that it is concerned with purity in distinction from the separated (+)-enantiomer. 76 For these reasons Apotex's molecule construction must be rejected. However, the rejection of this construction does not mean that claim 1 claims 100% pure separated (+)-enantiomer (as Sandoz would have it). It claims simply the separated (+)-enantiomer. 77 Apotex (and thus Sandoz) claimed that there is a fundamental contradiction between the fact that the term of the 144 patent was extended (which required the conclusion that Cipramil contained the substance, under s 70(3)(a) of the Patents Act) and the Full Court's construction of claim 1 in FFC#1 and FFC#3 that "reads '(+)-citalopram' as defining something that is 'separated, isolated or pure' and 'not in a racemate or mixture' with (-)-citalopram". Leaving aside the fact that these submissions depended on giving a meaning to "separated, isolated or pure" in the Full Court decisions which they do not bear, the purpose of this submission is not apparent. To the extent it is intended to support the proposition that the Full Court's construction of claim 1 is wrong, I do not see the contradiction. The fact that Cipramil contains the (+)-enantiomer within the meaning of s 70(3)(a) is a judicially determined fact, not now open to question. It is a fact, moreover, which accords with the evidence in the present case. While Professor Davies' evidence about the nature of the racemic mixture included concepts that I can best describe as involving an appreciation of chemical ontology far beyond that which would be attributed to the skilled addressee, nothing he said (or could have said), undermines the unassailable conclusion that founded the grant of the extension of term of the 144 patent. The construction of claim 1 which has prevailed in all previous cases, and which I also accept, is not inconsistent with this conclusion. The asserted inconsistency is a construct of Apotex's making, which pre-supposes both the correctness of its molecule construction and that the claimed (+)-enantiomer carries with it the concept of purity when it does not. All that has ever been said by previous cases, of current relevance, is that it is inherent within the very concept of the (+)-enantiomer that it is separated. This has never carried the freight which Apotex sought to attach to it.

3.3 Clarity 78 Apotex proposed that if claim 1 is construed as the Full Court in FFC#1 and FFC#3 would have it, and without any limitation as to purity, then the claim must fail for lack of clarity because the boundaries of claim 1 cannot be ascertained. It cannot be known what level of purity would infringe claim 1 because there is no apparent standard by which the claim can be interpreted. The 95% purity standard which Professor Davies proposed, and which Lindgren J accepted, was rejected in FFC#1 at [157]-[160] and is not advanced by Lundbeck in the present case. In any event, whether the proposed 95% purity is chemical, enantiomeric, optical or isomeric is unknown as Bennett J noted at [127] of FFC#1. 79 Sandoz, I note, did not adopt this aspect of Apotex's argument or, at the least, did not do so for the purpose of asserting that claims 1 and 3 of the 144 patent are invalid. To ensure that my reasoning for rejecting Sandoz's construction argument is properly exposed I deal with the clarity issues which Apotex raised below. 80 According to Apotex Bennett J's observation in FFC#1 at [159] was prophetic. Bennett J said: The construction contended for by Lundbeck also raises questions of clarity and uncertainty of infringement. 81 Her Honour was not referring to her preferred construction, however. She was referring to Lundbeck's then proposed construction that involved a limitation on the level of purity of 95%. Bennett J (and Middleton J) rejected this limitation because the claim was to the separated enantiomer, the specification referred not to 95% enantiomeric purity (Lundbeck's then case), but 99.6% and 99.9% optical purity having been achieved in Example 1. Her Honour was not saying that her own preferred construction involved a lack of clarity. Nor was she saying that the claim meant the separated (+)-enantiomer at 100% purity of any kind. 82 Apotex said that this was not a mere puzzle at the edge of a claim, as referred to in General Tire & Rubber Co v Firestone Tyre and Rubber Co [1972] RPC 457 at 515. The skilled addressee, in the present case, cannot determine where the edge lies but, for the claim to be clear, the skilled addressee must know how much (−)-enantiomer would be too much and infringe the claim. 83 Contrary to these submissions, I am persuaded that this is precisely the kind of matter to which the observations in General Tire at 515-516 apply. It was there said that: It is clear in our judgment that the question whether the patentee has sufficiently defined the scope of his claims is to be considered in relation to the facts of each case, that allowance is to be made for any difficulties to which the circumstances give rise, and that all that is required of the patentee is to give as clear a definition as the subject matter admits of. It is also clear in our judgment that, while the court is to have regard to all the relevant facts, the issue of definition is to be considered as a practical matter and little weight is to be given to puzzles set out at the edge of the claim which would not as a practical matter cause difficulty to a manufacturer wishing to satisfy himself that he is not infringing the patent. We accept also that definition of the scope of a claim is not necessarily insufficient because cases may arise in which it is difficult to decide whether there has been infringement or not provided the question can be formulated which the court has to answer in the issue of infringement. 84 As Lundbeck also noted, in Populin v HB Nominees Pty Ltd [1982] FCA 37; (1982) 41 ALR 471 at 476, the Full Court (Bowen CJ, Deane and Ellicott JJ) said this: The essential features of the product or process for which it claims a monopoly are to be determined not as a matter of abstract uninformed construction but by a common sense assessment of what the words used convey in the context of then-existing published knowledge. 85 Approaching the present case as a practical matter, it must be accepted that it was common ground for those skilled in this area of stereochemical synthesis that 100% purity of any kind was not achievable. None of the experts suggested that they did not know that the (+)-enantiomer meant the separated (+)-enantiomer. Professor Davies and Dr Robertson explained that in order to obtain the (+)-enantiomer at some level of purity less than the very high levels disclosed in both the specification and, for that matter, the generic parties' own batch analyses of their products, it would be necessary to add back in the racemic mixture. The commercial or practical purpose of such an exercise is not apparent. As noted, Dr Robertson described this as perverse, as it is. 86 Approached from a common sense perspective, in light of the knowledge which would be attributed to the skilled addressee of the 144 patent, the claim is clear. It is a claim to the separated (+)-enantiomer. As such, the racemic mixture, citalopram, falls outside the scope of the claim. As the skilled addressee would know the separated (+)-enantiomer cannot exist in a form that is 100% pure. There will always be trace, trivial or insignificant amounts of the (−)-enantiomer present. It does not matter what descriptor is used. The fact is that, to the person skilled in the art, the presence of insignificant amounts of the (−)-enantiomer is immaterial. Neither Professor Davies nor Dr Robertson (nor, for that matter, Dr Oppenheim once he put to one side what he incorrectly understood FFC#1 to mean) had any difficulty in identifying the separated (+)-enantiomer. They did not need to be instructed about any specific percentage tolerance to know when they were dealing with the (+)-enantiomer in contrast to the racemate. 87 Apotex's attempt to blur the boundaries of the claim by referring to Professor Davies' evidence about 95% purity is unsustainable. The fact that Professor Davies would read the claim as permitting up to 5% enantiomeric impurity (a proposition with which Dr Robertson agreed in oral evidence) does not render the claim unclear. This qualification was rightly rejected in FFC#1. Further, it was clear from the evidence of Professor Davies and Dr Robertson as a whole that they did not consider any specific percentage qualification was essential in order to understand the claim. They addressed the issue because they had been instructed to do so as a result of the dispute between the parties about what "separated, isolated or pure", as used in FFC#1, meant. This question, of what the majority in FFC#1 meant by this reference, was nothing more than a distraction. As discussed, it is apparent from FFC#1 that the majority were not saying that the (+)-enantiomer had to be 100% pure. But nor were they saying that, as a result, the (+)-enantiomer as claimed was the same as the racemate. Nor were they saying that claim 1, as construed in FFC#1, was or would be unclear. 88 Once this is accepted, it is apparent that Apotex's (and thus Sandoz's) recourse to various percentages of (−)-enantiomer, such as 96%, 95% or 94%, involves the creation of a hypothetical construct divorced from the reality of the relevant science. The skilled addressee would not have any difficulty in knowing when the claim was infringed. If the act involves the (+)-enantiomer, the acts are within the monopoly claimed.

3.4 Novelty 89 Again, Sandoz did not claim that the 144 patent was invalid for lack of novelty, but explaining Apotex's arguments and why I reject them may assist in understanding why I also reject Sandoz's construction and non-infringement arguments. 90 Apotex's novelty case may be shortly stated. Its principal case was that if its molecule construction is accepted, there is no question that claim 1 (and dependent claims) of the 144 patent will lack novelty because of the 445 or Citalopram patent. Its alternative case was that if the claimed invention in claim 1 of the 144 patent is merely the separated (+)-enantiomer then the 445 patent, nevertheless, deprives the claimed invention in claim 1 of the 144 patent of novelty. 91 Apotex submitted that the fact that the 445 patent does not refer to citalopram as having two enantiomers, the (+)-enantiomer and the (−)-enantiomer is irrelevant. The 445 patent disclosed the structural formula of citalopram which, to the skilled addressee, also disclosed that it was a racemic mixture consisting of equal amounts of the (+)-enantiomer and the (−)-enantiomer. The only matter not disclosed in the 445 patent, submitted Apotex, is the method of separating the two enantiomers, which is disclosed in and the subject of claim 6 of the 144 patent. 92 According to Apotex: (1) "[w]ith respect to Lindgren J, and the majority in the 2009 Full Court, it is an error to ask whether the prior art contains a direction to the skilled addressee to resolve a racemate into enantiomers, or discloses a means of doing so…"; and (2) "[t]he acceptance by Lindgren J (at [171]) and Bennett J (FFC #1 at [193]), that 'the skilled but non-inventive addressee would have understood that [racemic] (±)-citalopram consisted of the (+)-enantiomer and the (-)-enantiomer…' concludes the inquiry as to the novelty of claim 1, if Apotex's … submissions on construction are accepted". 93 Lundbeck's answer to Apotex's novelty challenge involves three aspects. 94 One, the issue of novelty was decided in the Lindgren J judgment and FFC#1, and there is no reason to depart from the conclusions there reached. The novelty challenge before Lindgren J was the same as that put by Apotex in these proceedings. It failed before Lindgren J and on appeal. Nothing has changed and the reasoning therein should be followed. 95 Two, in any event, Apotex is estopped from challenging the novelty of the 144 patent because it was a party to FFC#3 and the proper construction of claim 1 of the 144 patent, on which Apotex's novelty challenge depends, was an essential aspect of the reasoning in FFC#3 on which the judgment depends. 96 Three, the evidence in the present case means that it is not open to conclude that claim 1 of the 144 patent is invalid because the claimed invention was not novel. 97 Lundbeck's second proposition may be rejected immediately. Apotex's novelty challenge does not depend on its molecule construction being accepted. Apotex also contends that, on the construction which has prevailed thus far and which I also accept, the 445 patent destroys the novelty of the invention claimed in claim 1 (and dependent claim 3) of the 144 patent. Accordingly, there is no relevant issue estoppel which arises from FFC#3. 98 Lundbeck's first and third propositions, however, are persuasive and I accept them. I do not accept that Lindgren J and the majority in FFC#1 erred in rejecting the challenge to the novelty of the invention claimed in claim 1 of the 144 patent. Apotex's arguments are the same as those Lindgren J rejected at [169]-[171] of the Lindgren J judgment. As Lindgren J said at [171]: The Australian Citalopram Patent did not refer to "enantiomers". It did not expressly or by implication otherwise disclose the individual enantiomers. It disclosed the racemate and enabled the obtaining of it. Whether this anticipated the present invention turns on my construction of the Patent - see Section C above. The skilled but non-inventive addressee reading the Australian Citalopram Patent would have understood that (±)-citalopram consisted of the (+)-enantiomer of citalopram and the (-)-enantiomer, each as to 50%, and would have been able to identify the formulae for the S and R enantiomers, but would not have known, in the absence of experimentation, which was (+) and which was (-). These facts would not, however, point specifically to the independent existence of the enantiomers which is, according to my construction of the Patent specification, of the essence of claim 1. If I had construed claim 1 as referring to (+)-citalopram when present in the unresolved racemate, the Australian Citalopram Patent would have been an anticipation. But because of my construction outlined at Section C above, a person taught by the Australian Citalopram Patent, although taught to desire to obtain the racemate, would not be taught to desire to obtain the specific (+)-enantiomer in its own right. 99 In FFC#1 Bennett J, with whom Middleton J agreed, said this: 193 The prior citalopram patent described the racemate. It did not describe the pure or isolated (+)-enantiomer. There is no anticipation unless the disclosure of the racemate was, to the skilled addressee, a disclosure of the (+)-enantiomer. As the primary judge pointed out at [171], the skilled but non-inventive addressee would have understood that (±)-citalopram consisted of the (+)-enantiomer and the (-)-enantiomer and would have been able to identify the formulae for the S and R enantiomers but would not have known in the absence of experimentation which was the (+)-enantiomer and which the (-)-enantiomer. As his Honour said, these facts would not point specifically to the independent existence of the enantiomers. They did not disclose an invention which, if performed, would necessarily infringe the Patent. 194 It is the case that the skilled addressee knew that the racemate could be resolved into the enantiomers but there was nothing to tell him or her to do so. Further, the prior citalopram patent was silent as to the means of obtaining the enantiomers and there were different methods available to try to do so. There were no clear and unmistakable directions to obtain the enantiomers. Some of the available methodology may have been successful, other methods may not. 195 The prior citalopram patent does not render claim 1 of the Patent not novel. It follows that it does not render claims 2, 3, 4 and 5 not novel. 100 To my mind, this reasoning involves no error. There being no material difference in the evidence in the present case about the knowledge to be attributed to the skilled addressee, there is no proper foundation for any different conclusion. Apotex appeared to accept that there is no direction of any kind to make the (+)-enantiomer. Its case is that, consistent with the alternative tests of clear direction or clear description in General Tire at 485-486, the (+)-enantiomer is clearly described in the 445 patent because the skilled addressee would know that the racemate includes equal parts of the (+)-enantiomer and (−)-enantiomer. But as Lindgren J and the majority in the Full Court reasoned, knowing that the racemate includes equal parts of the (+)-enantiomer and (−)-enantiomer but not knowing which was the (+)-enantiomer and which the (-)-enantiomer, how to obtain either one, or why anyone might wish to obtain either one is hardly a clear description of the (+)-enantiomer which could be said to have made the claimed invention "apparent to the skilled addressee", as referred to be Bennett J in FFC# 1 citing Nicaro Holdings Pty Ltd v Martin Engineering Company [1990] FCA 37; (1990) 91 ALR 513 at 529. Nor did the 445 patent, as Apotex argued, make known the invention claimed in the 144 patent or enable that invention to be "read out of the prior publication", as referred to in Hill v Evans (1862) 1A IPR 1 at 6. Nor, from the 445 patent, would the skilled addressee clearly infer the existence of the claimed invention in claim 1 of the 144 patent, being the separated (+)-enantiomer, as referred to by Bennett J in FFC#1 at [169] and [171], citing in support C Van der Lely NV v Bamfords Ltd [1963] RPC 61. 101 It is for these reasons that, having identified the relevant principles, Bennett J rejected the appeal on the lack of novelty ground. Yet Apotex's argument involves accepting that while her Honour identified the very same principles upon which Apotex relied, her Honour nevertheless failed to apply those principles. The reality is that, properly applied, the principles simply do not support the asserted lack of novelty of the invention in claim 1 of the 144 patent. 102 Apart from this, I accept Lundbeck's submission that the expert evidence in the present case does not support the asserted lack of novelty. Adopting Lundbeck's submissions in this regard, Professor Davies said that the 445 patent did not disclose the invention claimed in claim 1 of the 144 patent because, so far as the 445 patent is concerned: i. the structures that are depicted are racemic; ii. there is no three dimensional information which discloses independently existing enantiomers; iii. the (+)-enantiomer is a different material from the racemate and has different properties to the racemate; and iv. there is nothing which tells him to resolve the racemate or how to resolve the racemate. 103 I also accept Professor Davies' evidence that figure 1 in the 445 patent does not describe individual enantiomers because it provides no stereochemical descriptor. While the existence of the two enantiomers could be conceived of as a "sort of paper exercise", they could not be identified and nothing could be known about them, not even whether they could be brought into existence separate from each other. Further the racemate and the (+)-enantiomer had to be understood to be "different materials", with "different physical properties", "different melting points", "different solubility rates" and, perhaps, different therapeutic benefits. 104 Dr Robertson was of the same view, noting the following matters about the 445 patent: v. the way the structure has been drawn (the bonds are drawn in the plane of the page, indicating a racemate); vi. the absence of reference to enantiomers; vii. the absence of prefixes indicating enantiomeric orientation (such as R/S or (+)/(−)); viii. the synthetic routes described, which will only yield racemic products; ix. the absence of a statement which instructs the reader to resolve the racemate, or suggests this has been done; and x. the absence of any information as to which enantiomer is the (+)-enantiomer, and which the (−)-enantiomer. 105 As Lundbeck submitted, consistently with this evidence: Formula 1 in the Citalopram Patent is immediately followed by the words: "It is an object of the present invention to provide the phthalans of formula 1, a method of making same …". The only methods of making "same" are methods of making racemic mixtures. In other words, by those words and the descriptions that follow the Citalopram Patent confirms that Formula 1 is only contemplating racemic mixtures. There is not a single word in the Citalopram Patent to suggest otherwise. 106 Dr Oppenheim's evidence, to the extent it diverged from the views of Professor Davies and Dr Robertson, was unpersuasive. Having heard the evidence of Professor Davies, Dr Robertson and Dr Oppenheim it was apparent that the latter's expertise as a formulator and not a medicinal chemist undermined the reliability of his opinions in a number of critical respects. In short, claim 1 of the 144 patent is not concerned with formulation. Some aspects of claim 3 may cover formulation, but they are immaterial for the purposes of the present dispute. The skilled addressee is therefore a medicinal chemist. Professor Davies and Dr Robertson are medicinal chemists. Dr Oppenheim is not. The difference in relevant expertise between Professor Davies and Dr Robertson (on the one hand) and Dr Oppenheim (on the other hand) was obvious in a number of critical respects. It was difficult to avoid the conclusion that Dr Oppenheim had misunderstood the reaction scheme described in the 144 patent in a manner so fundamental that his descriptions of it were simply nonsensical to Professor Davies and Dr Robertson. This was not merely a careless description or sloppy work, as Dr Oppenheim would have it. It exposed his lack of expertise in the relevant area of discourse. Indeed, Dr Oppenheim's only experience in resolving enantiomers was as an undergraduate in the 1960s and 1970s. 107 Given that I do not consider Dr Oppenheim's evidence to be reliable given his lack of relevant expertise it is not necessary to decide why Dr Oppenheim agreed in the joint report with Professor Davies and Dr Robertson that the 445 patent does not contemplate a single enantiomer, and then in oral evidence sought to resile from this proposition. Dr Oppenheim, ultimately, could not give any cogent explanation as to why, in his oral evidence, he departed from his usual practice of reading a compound without a stereochemical descriptor (as in the 445 patent) as a reference to a racemate, which would be consistent with the opinions he gave in the joint report. In any event, this part of Dr Oppenheim's evidence went nowhere. As Lundbeck submitted: Dr Oppenheim's evidence under cross-examination was that the formula in the citalopram patent could represent something else, but he accepted that having regard to the context and what followed, it only referred to the racemic mixture: "That is how they have used it". He also volunteered that his reading of the diagram as possibly referring to the enantiomers was in the absence of any context: "I'm drawing a distinction between what formula 1 represents to me and also the rest of the [Citalopram] patent. 108 Apotex's novelty challenge, accordingly, must be rejected.

3.5 Fair basis 109 I adopt the same approach to Apotex's fair basis argument, that is, I deal with the argument for the purpose of ensuring that my reasoning about Sandoz's construction argument is exposed. 110 Apotex's contention that claim 1 (and dependent claim 3) is not fairly based on matter described in the specification is based on the same artificial construct rejected above in relation to the alleged lack of clarity of claim 1 of the 144 patent. 111 Apotex's argument was that none of the examples in the 144 patent disclose the (+)-enantiomer in the absence of the (−)-enantiomer and, as such, there is no real and reasonably clear disclosure of the separated, isolated or pure (+)-enantiomer. The argument must be rejected for the same reasons that the lack of clarity argument must be rejected. The 144 patent discloses the separated (+)-enantiomer. The fact that the examples also disclose the scientific inevitable presence of trace, trivial or insignificant amounts of the (−)-enantiomer is irrelevant. That scientific inevitability is inherent within the very concept of the separated (+)-enantiomer.

3.6 Conclusions about validity 112 For the reasons given, Apotex's challenge to the validity of claims 1 and 3 of the 144 patent must be rejected. For the same reasons, Sandoz's construction argument must be rejected.

Infringement of the 144 patent 113 Given the proper construction of claim 1 of the 144 patent, it is apparent that Sandoz's defence to infringement based on the presence of trace, trivial or insignificant amounts of the (−)-enantiomer in its products cannot succeed. The batch records for the allegedly infringing products show that the escitalopram oxalate products of Sandoz's products as sold in Australia between 13 June 2009 and 9 December 2012, as was scientifically inevitable, contained trace, trivial or insignificant amounts of the (−)-enantiomer. For Sandoz, the batch records show the separated (+)-enantiomer at chiral purity levels varying between 99.4% and 99.8%. These are mere trace, trivial or insignificant amounts which show that the active ingredient in the products is the separated (+)-enantiomer as claimed in claim 1 of the 144 patent. 114 As discussed, this does not mean the products did not infringe claim 1 as such trace, trivial or insignificant amounts of the (−)-enantiomer are inherent within the concept of the separated (+)-enantiomer claimed. Considerations of fitness for purpose are beside the point. Sandoz sold escitalopram oxalate products which is the separated (+)-enantiomer. There can be no defence to infringement based on the presence of trace, trivial or insignificant amounts of the (−)-enantiomer as disclosed in the batch records. 115 Accordingly, leaving aside the various other arguments about the operation of the Patents Act, the status of Lundbeck AU as an exclusive licensee or not, and Sandoz's claim that it had a licence authorising it to exploit the invention in the period of the alleged infringement, I am satisfied that Sandoz infringed claims 1 and 3 of the 144 patent.

Section 78 116 Section 78 of the Patents Act provides that: If the Commissioner grants an extension of the term of a standard patent, the exclusive rights of the patentee during the term of the extension are not infringed: … (b) by a person exploiting any form of the invention other than: (i) a pharmaceutical substance per se that is in substance disclosed in the complete specification of the patent and in substance falls within the scope of the claim or claims of that specification; … 117 Sandoz contended that the presence of the (−)-enantiomer in their products means that it was not exploiting a pharmaceutical substance per se as required by s 78(b)(i) of the Patents Act. According to this argument, the effect of the reasoning in FFC#3 supports this contention. Sandoz pointed to the reasons given in FFC#3 at [38]-[41], [46], [93] and [96] in particular, recalling that the issue in FFC#3 concerned s 70(2)(a) of the Patents Act which enabled the term of the 144 patent to be extended only if it satisfied the requirement that "[o]ne or more pharmaceutical substances per se, must in substance be disclosed in the complete specification of the patent and in substance fall within the scope of the claim or claims of that specification". In FFC#3 the Full Court, referring to FFC#1 as the Decision, said: 93 Semantics aside, it is clear that the claim describes a pharmaceutical substance per se. The substance was, as explained by Lindgren J at [536], a new chemical entity. The racemate and the (+)-enantiomer had different physico-chemical interactions manifested in different pharmacodynamics and pharmacokinetics (the Decision at [234]). The claim is to (+)-citalopram, irrespective of how it is produced. The isolated (+)-enantiomer plainly qualifies as a pharmaceutical substance per se and the primary Judge was correct in concluding that it satisfies s 70(2)(a) of the Act. 94 As pointed out by Lundbeck, Alphapharm's submission that Bennett and Middleton JJ construed the claim to include an additional integer or a limitation, such that the claim was no longer to the (+)-enantiomer as a pharmaceutical substance per se, is wholly inconsistent with their Honours' conclusion that the subject matter of the claim satisfied s 70(2)(a) of the Act, such that Lundbeck was entitled to an extension of term based upon Cipramil but not on Lexapro. 96 It is clear from the Decision that the expressions such as "separated or isolated or pure" did not import any concept of degree of purity. Quite clearly, Bennett J used the expression as a hendiadys, to explain that the (+)-enantiomer alone was claimed - without the (−)-enantiomer or the racemate. That is, it was the enantiomer itself. Put another way, it was the enantiomer per se. 118 Sandoz also referred to [40] of FFC#3 in which the decision of the delegate of the Commissioner was considered. The delegate's reasoning was described as follows: 39 The Delegate's reasoning was straightforward: • Bennett and Middleton JJ concluded that the reference to (+)-citalopram in the claim without more indicates that it is the separated or isolated or pure enantiomer that is claimed. • "Per se" means taken alone 'essentially without reference to anything else' (Pharmacia Italia SpA v Mayne Pharma Pty Ltd (2006) 69 IPR 1 at [94]). • By definition a compound is a "substance" alone. As a pure compound it must be a pharmaceutical substance per se. • Pure escitalopram is read as free of the negative enantiomer. If the pharmaceutical substance is viewed as pure citalopram, the claim is directed to the pharmaceutical substance and nothing more than the pharmaceutical substance. 40 This was considered with and supported by the evidence of [Dr] Oppenheim at [163], where he stated that the molecule must be 100% pure if it is separated and isolated. 41 The Delegate accepted that escitalopram in its pure form was not used in the preparation of Cipramil but also that Cipramil has within it molecules of escitalopram. He noted that in the Lundbeck Full Court, Bennett J concluded that, given that the racemic mixture includes both the (−)-enantiomer and the (+)-enantiomer, Cipramil must "contain" the (+)-enantiomer. The Delegate agreed that this approach was correct. 119 Sandoz's case on s 78 culminated in this submission: …Although FFC #3 also found that citalopram contained the pharmaceutical substance (+)-citalopram for the purpose of s 70(3), it could not have found that citalopram (a racemic mixture) contained "(+)-citalopram in the absence of (-)-citalopram" or "(+)-citalopram per se". Put another way, FFC #3: (a) identified (+)-citalopram (whatever its surrounds, including as part of a racemic mixture) as the pharmaceutical substance for the purpose of s 70(3); and (b) concluded that (+)-citalopram per se (in the absence of (-)-citalopram fell within the scope of claim 1). If this Court adopts a construction pursuant to which the Products are found to fall within the scope of claim 1 despite the fact that they contain (-)-citalopram, the Patent is nonetheless not infringed by dealings in those products during the extended term (by operation of s 78(b)(i)). That is because those products (which, like CIPRAMIL, contain (-)-citalopram) must, by reason of FFC #3's conclusions, be a form of the invention other than the pharmaceutical substance (+)-citalopram per se. They do not contain (+)-citalopram in the absence of (-)-citalopram. 120 These arguments involve the same artificial construct as those rejected above. The pharmaceutical substance (+)-citalopram per se is the separate (+)-enantiomer. It cannot exist in a form which is 100% free from the (−)-enantiomer. The fact that it is a scientific inevitability that the separate (+)-enantiomer will not exist in a form which is 100% free from the (−)-enantiomer is a part of the inherent character of that pharmaceutical substance per se. It necessarily follows from the reasoning above that the generic parties are exploiting a form of the invention in which a pharmaceutical substance per se that is in substance disclosed in the complete specification of the patent and in substance falls within the scope of the claim or claims of that specification. As a result, the exception to non-infringement in s 78(b)(i) is engaged. Sandoz cannot rely on s 78 to avoid infringement of the 144 patent. 121 One other observation should be made, because it is relevant to another issue requiring resolution, which is the operation of s 79 of the Patents Act. It will be apparent that Sandoz assumed that its argument, if accepted, would prevent both Lundbeck as the patentee and Lundbeck AU (if it was the exclusive licensee of the 144 patent, an issue resolved against the generic parties below) from claiming infringement. Sandoz assumed this despite the fact that s 78 says that "the exclusive rights of the patentee during the term of the extension are not infringed". The assumption is correct, as I explain below in the context of s 79, as it properly reflects the overall scheme of the legislation. By operating on the exclusive rights of the patentee during the extended term of the patent, s 78 also has effect with respect to the rights of an exclusive licensee.

6.1 The issue 122 Sandoz contended that s 223(10) of the Patents Act prevents Lundbeck from bringing these proceedings because the 144 patent "ceased" on 13 June 2009 by reason of expiry of its 20 year term and was "restored" on 25 June 2014 when the Commissioner granted the extension of term of the 144 patent. Lundbeck contended that s 223(10) does not apply because the 144 patent did not "cease" and was not "restored" as provided for in s 223(10). Rather, the 144 patent expired on 13 June 2009 but, because its term was extended to 9 December 2012 on 25 June 2014, s 79 applies to infringement proceedings relating to the period between 13 June 2009 and 9 December 2012. 123 I am persuaded that Lundbeck's approach is correct.

6.2 Statutory provisions 124 Section 223(10) is in Chapter 22 of the Patents Act which contains miscellaneous provisions. Section 223 is in these terms: (1) The Commissioner must extend the time for doing a relevant act that is required to be done within a certain time if the act is not, or cannot be, done within that time because of an error or omission by: (a) the Commissioner or a Deputy Commissioner; or (b) an employee; or (ba) a New Zealand delegate; or (c) a person providing, or proposing to provide, services for the benefit of the Patent Office; or (d) the receiving Office; or (e) the International Bureau of the World Intellectual Property Organization. (1A) For the purposes of subsection (1), it is immaterial whether a relevant act took place, or is to take place, in New Zealand. (1B) For the purposes of subsection (1), it is immaterial whether an error or omission took place in New Zealand. (2) Where, because of: (a) an error or omission by the person concerned or by his or her agent or attorney; or (b) circumstances beyond the control of the person concerned; a relevant act that is required to be done within a certain time is not, or cannot be, done within that time, the Commissioner may, on application made by the person concerned in accordance with the regulations, extend the time for doing the act. (2A) If: (a) a relevant act that is required to be done within a certain time is not done within that time; and (b) the Commissioner is satisfied, on the balance of probabilities, that the person concerned took due care, as required in the circumstances, to ensure the doing of the act within that time; the Commissioner must, on application made by the person concerned in accordance with the regulations and within the prescribed period, extend the time for doing the act. (2B) An extension of time under subsection (2A) cannot exceed the period prescribed for the purposes of this subsection. (3) The time allowed for doing a relevant act may be extended, whether before or after that time has expired. (3A) Despite subsection (3), the time allowed for doing a relevant act may be extended under subsection (2A) only after that time has expired. (4) The Commissioner must advertise in the Official Journal : (a) an application made for an extension of time for more than 3 months; or (b) an application made for an extension of time for doing a prescribed relevant act in prescribed circumstances. (6) Subject to subsection (6A), a person may, as prescribed, oppose the granting under subsection (2) or (2A) of the application. (6A) If the Commissioner is satisfied, on the balance of probabilities, that an application under subsection (2) or (2A) would not be granted even in the absence of opposition under subsection (6): (a) the Commissioner need not advertise the application in accordance with subsection (4); and (b) the application cannot be opposed, despite subsection (6); and (c) the Commissioner must refuse to grant the application. (7) Where: (a) a patent application lapses, or a patent ceases, because of a failure to do one or more relevant acts within the time allowed; and (b) the time for doing that act or those acts is extended; the application or patent must be treated as having been restored. (8) Where: (a) a provisional patent application lapses under subsection 142(1) at the end of the period prescribed for the purposes of section 38; and (b) that period is extended; the application must be treated as if it had not lapsed. (9) Where the Commissioner grants: (a) an extension of more than 3 months for doing a relevant act; or (b) an extension of time for doing a prescribed relevant act in prescribed circumstances; the prescribed provisions have effect for the protection or compensation of persons who, before the day on which the application for extension of time is advertised under subsection (4), exploited (or took definite steps by way of contract or otherwise to exploit) the invention concerned because of the failure to do the relevant act within the time allowed, the lapsing of the patent application or the ceasing of the patent, as the case may be. (10) Infringement proceedings cannot be brought in respect of an infringement committed: (a) between the day on which the patent application lapses and the day on which it is restored; or (b) between the day on which the patent ceases and the day on which it is restored. (11) In this section: "relevant act" means an action (other than a prescribed action) in relation to a patent, a patent application, or any proceedings under this Act (other than court proceedings), and includes the making of a Convention application within the time allowed for making such applications. 125 Other provisions of the Patents Act are relevant to the construction of s 223(10). 126 Chapter 6 of the Patents Act is entitled "Grant and Term of Patents". The date of a patent is regulated by s 65 which is in Pt 1 of Ch 6. The provisions regulating the term of patents are in Pt 2 of Ch 6. As noted, patents have a term which is fixed by s 67 for a standard patent (20 years from the date of the patent) and s 68 for an innovation patent (8 years from the date of the patent). Part 3 of Ch 6 deals with the extension of term of standard patents dealing with pharmaceutical substances. For present purposes it is sufficient to note that these provisions include s 70 which permits the patentee of a standard patent to apply for an extension of term of the patent if certain conditions are satisfied concerning pharmaceutical substances. Section 71(2), to which Lindgren J referred in [2006] FCA 163 at [13] (quoted above), regulates the timing of an application for an extension of term of a standard patent. Section 72 requires the Commissioner to publish notice of the application for the extension of term. Section 74 requires the Commissioner to accept or refuse the application depending on whether the Commissioner is or is not satisfied, on the balance of probabilities, that the requirements of ss 70 and 71 are satisfied in relation to the application. Section 75 permits any person to oppose the grant of an extension of term of a standard patent only on the ground that the requirements of ss 70 and 71 are not satisfied in relation to the application. Section 76 is in these terms: (1) The Commissioner must grant an extension of the term of a standard patent if: (a) there is no opposition to the grant; or (b) in spite of opposition, the Commissioner's decision, or the decision on appeal, is that the extension should be granted. (2) If the Commissioner grants an extension, the Commissioner must notify the applicant in writing of the grant and publish a notice of the grant in the Official Journal. 127 Section 77 regulates the length of the extension of term that may be granted. Section 78 confines the rights of the patentee during the extended term by providing that the exclusive rights of the patentee during the term of the extension are not infringed, relevantly: … (b) by a person exploiting any form of the invention other than: (i) a pharmaceutical substance per se that is in substance disclosed in the complete specification of the patent and in substance falls within the scope of the claim or claims of that specification; … 128 Section 79 is set out below in that part of this judgment dealing with arguments about its operation. 129 Sections 143 and 143A, referred to by the parties in their competing submissions, are as follows: 143 A standard patent ceases if the patentee: (a) does not pay a renewal fee for the patent within the prescribed period; or (b) does not file the prescribed documents (if any) within the prescribed period. 143A An innovation patent ceases if: (a) the fee for filing the request and accompanying specification relating to an application for an innovation patent is not paid in accordance with the regulations; or (b) after an examination of the patent has been requested under paragraph 101A(b), the patentee does not pay the prescribed fee for the examination within the prescribed period; or (c) the Commissioner does not make a decision under paragraph 101E(1)(a) within the period prescribed for the purposes of this paragraph; or (d) the patentee does not pay a renewal fee for the patent within the prescribed period; or (e) the patentee does not comply with a direction of the Commissioner under section 106 within the time allowed by the Commissioner under that section. 130 Other provisions identified as relevant to construction of s 223(10) include these: 55 … (2) Where a notice is published under paragraph 49(5)(b) in relation to an application for a standard patent, or under subsection 62(2) in relation to the grant of an innovation patent, the following documents are open to public inspection: … (b) all documents (other than prescribed documents) filed, after the patent ceases, expires or is revoked, in relation to the former patent; … 79C … (2) The patentee may make the further complete application only during the period: (a) starting when an examination of the first patent begins; and (b) ending when any of the following happens: (i) the term of the first patent ends; (ii) the first patent is revoked; (iii) the first patent ceases; (iv) a period prescribed by the regulations for the purposes of this subparagraph ends. 101E (1) This section applies to an innovation patent if: … (b) the patent has not ceased under section 143A. 101F (1) The Commissioner must revoke a patent if: … (c) the patent has not ceased under section 143A. 145 (1) A contract relating to the lease of, or a licence to exploit, a patented invention may be terminated by either party, on giving 3 months' notice in writing to the other party, at any time after the patent, or all the patents, by which the invention was protected at the time the contract was made, have ceased to be in force. 131 Other provisions also refer to variations of "cease", albeit in contexts other than a patent ceasing, being ss 41(4)(a), 42(1)(c), 133(6), 136H(2)(a)(i), 136J(2)(b), 144(5), 148(1)(d), 149, 151(1), 165A(2) and 215(3). "Unexpired" appears in s 85(2). Variants of "restore" appear in ss 150, 228(2)(ha)(iv) and 228(2)(j)(ii) and of "lapse" appear in ss 32(2), 36(2), 54(3), 142, 147(3)(c), 148, 149, 150, 227(4), 228(2)(ha)(iii) and (iv), 228(2)(i)(iii) and 228(j)(i) and (ii). These provisions need not be set out. All that need additionally be noted is that a patent application may lapse (ss 142 and 148) and if it lapses under s 148 may be restored under s 150.

6.3 Sandoz's submissions 132 Sandoz contended that s 223(10)(b) is engaged in the present case. The "day on which the [144] cease[d]", they say, was 13 June 2009. The day on which the 144 patent was "restored", they say, was 25 June 2014. Accordingly, Sandoz said, s 223(10) operates so that infringement proceedings cannot be brought for an infringement between, relevantly, 13 June 2009 and 9 December 2012, the latter being the date on which the 144 patent expired after extension of the term. 133 Sandoz submitted that this conclusion is supported by the following: (1) Section 223 is a remedial provision which confers a power to extend the time limits imposed by the Patents Act and, as a quid pro quo, provides protections for third parties which may be adversely affected by the extension. (2) It is common ground that s 223(2) and (10) would apply, for example, if a patent ceased because of a failure to pay a renewal fee as required by ss 143 or 143A of the Patents Act. (3) It was held in Alphapharm Pty Ltd v H Lundbeck A/S [2014] HCA 42; (2014) 254 CLR 247 that s 223(2) applies to a case involving an application for an extension of time to apply for an extension of term of a patent. Given the remedial nature of s 223 the Court should be slow to construe the section so that s 223(2) applies but s 223(10) does not. (4) There is no policy reason why third parties are protected by s 223(10) if a patent ceases for failure to pay a renewal fee but are not protected if time is extended to permit a patentee to apply for an extension of term. (5) As provisions have been added to the Patents Act, such as s 143A in 2000 concerning the ceasing of innovation patents, s 223(10) has correspondingly expanded to apply to those examples of ceasing of a patent. Further, s 143A is not confined to an innovation patent ceasing for failure to pay a renewal fee (see s 143A(c) and (e)). (6) The words "ceases" and "restore" are ordinary English words that are used in different manners throughout the Patents Act and are sufficiently broad to encompass the expiry of a patent due to the end of its term being reached and the extension of the term of the patent after that expiry date. In such a case the patent has ceased and is then restored. In contrast, if the term is extended before the patent reaches the end of its term then the patent will not cease and will not be restored and s 223(10) has no role to play. (7) The patent in s 223(10) is the patent in relation to which the extension of time is sought in s 223(2) to do a "relevant act" as defined in s 223(11). As a result, if an extension of term is filed within time, s 223(10) cannot apply. It is only if an extension of time is required to apply for an extension of term that s 223(10) may be engaged. (8) Section 79 does not override s 223(10). If an extension of time is required, then s 223(10) is engaged (see the reasoning of Beach J discussed below). Thus: …where the Patent would never have been restored but for the grant of an extension of time by the Commissioner, Lundbeck DN's "right to start infringement proceedings" remained subject to the s 223(10) bar. The "same rights" cannot be transformed into "superior rights" by s 79, nor can s 79 somehow "override" the consequences of s 223. (9) Section 223(7) is not an answer. Section 223(7) does not apply in the present case as it concerns a patent ceasing because of a failure to do one or more relevant acts within the time allowed. Section 223(7), by contrast, would apply to a case where fees had not been paid as required. Without s 223(7), there would be no restoration on payment of the fees before the extended time for payment. (10) In contrast to s 223(7), s 223(10) is not confined to cases where the patent ceases because of a failure to do one or more relevant acts within the time allowed. Without that qualification, the word "ceases" in s 223(10) includes a patent ceasing due to expiry of the patent. Further, for a patent which is the subject of the grant of an extension of term, restoration of the patent is effected by ss 76 and 77. (11) In the alternative, by operation of s 143(b), the 144 patent ceased on 13 June 2009 because Lundbeck A/S failed to do a relevant act within the time allowed (to file an application for an extension of term based on Cipramil within the time allowed for that to occur). It would then follow that s 223(7) applies and the 144 patent would be treated as if it had been restored no earlier than 1 June 2011 when the Commissioner granted the extension of time to apply for the extension of term. As, however, s 223(7) does not identify the date on which the patent is to be treated as having been restored the better view is that the patent is to be treated as having been restored on 25 June 2014 when the extension of term was granted. 134 Sandoz noted that this approach accords with Beach J's conclusions in H Lundbeck A/S v Commissioner of Patents [2017] FCA 56; (2017) 249 FCR 41. His Honour, in rejecting Lundbeck A/S's argument that s 223(9) did not apply, said: 109 First, in my view Lundbeck confuses the two different regimes. Sections 70 to 79A are dealing with the extension of term. Contrastingly, s 223 is dealing with extensions of time. And section 223(9) is dealing with the consequences of an extension of time. 110 Second, the regimes can work harmoniously together. For example, as the High Court discussed in Alphapharm, the second time requirement in s 71(2) is not excluded from s 223(2)(a) by reg 22.11(4)(b). Accordingly, s 223(2) applied to extend the time requirement calculated by reference to ss 71(2)(a), 71(2)(b) and 71(2)(c). 111 Third, the regimes also work harmoniously together by reason of the fact that the extension of time regime is anterior to the extension of term regime. In other words, until the extension of time has been validly made or given, there is no extension of term validly made or given. In other words, the regime in s 223 for the extension of time is anterior in one sense to the extension of term. Sections 223(2), 223(4), 223(7) and 223(9) have independent work to do. The present case does not concern the context of a general regime being inconsistent with a specific regime. 112 Fourth, Lundbeck says that s 79 is inconsistent with s 223(9). But they are dealing with different matters as I have said above. 135 Beach J also said: 130 First, the word "ceasing" is an ordinary English word. Its natural meaning encompasses "expiry". Now "ceasing" may have a narrower operation in some parts of the Act (such as s 55(2)(b)), but it is capable in other provisions of the Act of having a broader operation. Such an operation is apparent from s 145(1), which reads "the patent, or all the patents … have ceased to be in force". This has been interpreted to include expiry of the term of a patent (Regency Media Pty Ltd v MPEG LA, LLC (2014) 231 FCR 588 at [13] and [35] per Bennett and Pagone JJ); see also Alphapharm at [68] where it was noted that one could have a lapse of a patent due to the expiration of the term. 131 Second, if s 223(9) is to be construed so that a licence applicant must show that there was either a lapsing or ceasing of a patent to enliven the Commissioner's power to grant a licence (Lundbeck's construction), then the context and purpose of the provision support "ceasing" being given a wide operation, and not being necessarily confined to the meaning of "ceases" in s 143 of the Act. In the context of s 223(9) and reg 22.21, "ceasing" refers to a patent ceasing to be in force, a concept which includes both "ceasing" in a narrow sense and expiry of the term of a patent. 132 Finally, if "ceasing" has a more confined operation in s 223(9), even on Lundbeck's case it includes a patent which "ceased" within the meaning of s 143. As the Generics have contended, it would seem to be arguable that the present circumstances satisfy the definition of "ceasing" in s 143 subject to the difficulty that I raised with counsel… 133 But I do not need to rule on this argument or elaborate further on the difficulties that I raised with counsel. For present purposes, my determination is that "ceasing" can include "expiry".

6.4 Discussion 136 For the reasons which follow I consider that the submissions of Lundbeck, that s 223(10) does not apply to these proceedings for infringement of the 144 patent, is to be preferred as better reflecting the scheme of the Patents Act as a whole than any of the constructions proposed during the hearing. 137 First, the primary position of Sandoz, that s 223(7) does not apply in the present case (a proposition with which Lundbeck agrees), exposes a tension in the construction for which Sandoz contended. Section 223 as a whole is concerned with extensions of time for doing a "relevant act" as defined by s 223(11), that is doing an action in relation to, relevantly, a patent. Accordingly, s 223(1), (2) and (2A) empower the Commissioner to extend time for "doing an act" where the act is not or cannot be done in the required time in one of the identified circumstances. In the present case, for example, Lundbeck A/S did not make an application for an extension of term of the 144 patent based on Cipramil within the time required for the doing of that act because of an error and thus could make an application to extend the time for doing that act relying on s 223(2)(a). 138 Section 223(7) is to be understood in this context. By s 223(1) to (2A) time may be extended if an act in relation to a patent application or a patent is not or cannot be done within the time required for the doing of the act. Section 223(7) recognises that, by other provisions of the Patents Act, if an act in relation to a patent application or a patent is not or cannot be done within the time required for the doing of the act, a patent application may lapse or a patent may cease. The "failure to do one or more of the relevant acts within the time allowed" means nothing more than that the act is not or cannot be done within the required time as provided for in s 223(1) to (2A). 139 In the present case, if, as Sandoz would have it, the 144 patent ceased on 13 June 2009, it is a strained construction of s 223(7) to say that the 144 patent ceased, but not because of a failure to do one or more acts within the relevant time. Yet, as the alternative submission for Sandoz recognised, by far the better view is that s 223(7) does not apply to a patent which has "ceased" because an application for an extension of term has not been made and granted before expiry of the term of the patent. This is by far the better view because if s 223(7) applies then the patent must be treated as having been restored. And despite the submissions for Sandoz to the contrary, it is apparent that s 223(7) requires the patent to be treated as having been restored when the time for doing the act (on this example, applying to extend the term of the patent) is extended. This follows from the word "where" at the start of s 223(7). It is on satisfaction of both conditions that the patent must be treated as having been restored. This consequence of purported restoration of the patent under s 223(7) on the grant of the extension of time, however, cannot be reconciled with the regime for an extension of term of a patent. The patent will be required to be treated as if it has been restored on the day when the extension of time to apply for an extension of term is granted even though, at that time, no extension of term has been granted. This, to my mind, is nonsensical. In the present case, the 144 patent would have been "restored" on and from (for some undefined period) 1 June 2011 (the date the extension of time was granted). 140 Further, if Sandoz is right that s 223(7) does not prescribe the date of restoration (which I do not accept) but leaves the time of restoration at large then there is no reason that the restoring of the patent (or treating the patent as restored) in the present case should not be to a date consistent with the position which would apply under s 79 (that is, back to the date of the alleged infringing acts). The result would then be that there is no period upon which s 223(10) could operate. Given the statutory scheme for extensions of term there would be no reason to decide that the 144 patent should be treated as restored only on 25 June 2014, as Sandoz would have it. 141 But this aspect of the discussion need not be taken further as I consider s 223(7) requires a patent to which the section applies to be treated as having been restored as soon as the extension of time is granted. I also consider Sandoz's alternative submission, that s 223(7) applies to the present case, to be untenable. Read in this way, the 144 patent would have been restored on 1 June 2011, when time was extended for Lundbeck A/S to apply for the extension of term. If so restored, then what function could the extension of term play? The 144 patent did not need to have its term extended because it was restored, for some period undefined by the terms of s 223. For these reasons, Sandoz's alternative submission is not open. 142 Accordingly, in my view, s 223(7) does not apply, but not for the reasons given by Sandoz. As discussed further below, Sandoz's principal construction also results in a statutory scheme which is incoherent. 143 Sandoz relied on the words in s 223(7) "because of a failure to do one or more relevant acts within the time allowed" to avoid the provision applying to an extension of time to apply for an extension of term of the patent where the extension of term is made during the term (as required by s 71(2)), but the extension of time and/or term is not granted during the (unextended) term of the patent. Sandoz's case suggested that "because of a failure to do one or more relevant acts within the time allowed" in s 223(7) must mean something more than simply an act required to be done that is not or cannot be done as provided for in s 223(1) to (2A) and the existence of some cause and effect relationship between that act not being done and the ceasing of the patent. Once it is accepted, as I consider it should be, that a "failure to do one or more relevant acts within the time allowed" in s 223(7) is simply an act that is not done or cannot be done within the time required as provided for in s 223(1) to (2A) which has some causal connection with the patent ceasing then, on the case of Sandoz, avoiding s 223(7) becomes near impossible. In short, on this approach, it would seem to fit with a natural and ordinary reading of s 223(7) that, to take this case as an example, the 144 patent ceased because of Lundbeck's failure to apply for an extension of term within the required time. This is the equivalent of the argument which Beach J considered in [2017] FCA 56 at [132] in relation to s 223(9). Section 223(9) gives effect to the prescribed provisions (relevantly, reg 22.21 of the Patents Regulations 1991 (Cth)). Under this part of the scheme, in the circumstances identified in s 223(9), the Commissioner may grant a licence to a person who exploited the invention (or took definite steps to exploit it) for the period of time extended under s 223(1), (2) or (2A) (see reg 22.21(2)(c)). The relevant circumstances, as Beach J noted in [2017] FCA 56 at [132], are not confined to the lapse of the patent application or the ceasing of the patent. They include, as an alternative, exploiting the invention "because of the failure to do the relevant act within the time allowed". The time periods in s 223(9) and (10) may overlap but I do not see that as indicating that any particular construction is to be preferred. 144 Second, and another indicator that scrutiny of the Patents Act as a whole is justified, is that if a patent ceases on expiry of its term and is restored on an extension of term being granted (assuming the grant of the extension of term after expiry) within the meaning of s 223(10) then, on first reading, s 223(10) would apply to any patent the term of which has been extended after the standard term has expired even if no extension of time has been granted. This is because, on Sandoz's case, such a patent will have ceased and been restored and s 223(10) simply refers to "infringement proceedings cannot be brought", not confining itself to infringement proceedings for a patent in respect of which time for doing an act has been extended under s 223(1), (2) or (2A). Sandoz also sought to avoid this result, which it recognised to be irreconcilable with the extension of term provisions including s 79, by stressing that, relevantly, "the patent" as referred to in s 223(10)(b) can only be the patent in relation to which a relevant act is not or cannot be done as provided for in s 223(1), (2) or (2A). I agree that this is a better reading but this again exposes the fact that the scheme of s 223 is founded on the capacity to extend time for acts that are not or cannot be done within the allowable time. Some of those acts, if not done or unable to be done, may cause a patent application or lapse or a patent to cease, which is where s 223(7) comes into play. The causal connection requirement in s 223(7) (because of a failure to do one or more relevant acts within the time allowed) must be understood in this context. The fact that the "relevant act" is not done or cannot be done within the time allowed need not be the sole or direct cause of the lapsing or ceasing for s 223(7) to be engaged. The requirement "because of a failure to do one or more relevant acts within the time allowed" in s 223(7), in context, is satisfied by a cause and effect connection between not doing a relevant act and lapse or cease. If this is so, Sandoz's attempt to escape s 223(7), because of its manifest inconsistency with its principal construction of s 223(10), is thus confounded. 145 The same considerations also expose the fine distinctions on which Sandoz's principal construction depends. As Sandoz would have it, "the patent" in s 223(10) is the patent in relation to which an act has not been or cannot be done (that is, a relevant act as per s 223(11)) as referred to in s 223(1), (2) or (2A), yet if that patent "ceases" by reason of expiry as its term has ended "the patent" is the patent as referred to in s 223(1), (2) or (2A) and s 223(10) (and, for that matter, s 223(9)), but is not a patent which "ceases because of a failure to do one or more relevant acts within the time allowed" as referred to in s 223(7). If a patent can cease because of expiry of its term and a relevant act can be not applying for an extension of term within time, then this is not a self-evidently natural reading of the whole of s 223 as a coherent scheme. It is apparent that one part of the scheme is based on s 223(7) providing that a patent application which has lapsed or a patent that has ceased is restored where (which is also to be understood, in context, as when) the time for doing the relevant act is extended and another part of the scheme is based on s 223(10), ensuring that for the period in between lapse or cease and restoration as provided for in s 223(7), infringement proceedings cannot be brought. 146 In other words, it is s 223(7) that creates the gap or the period between lapsing or ceasing and restoring upon which s 223(10) can operate. Given this, it would be odd if the scope of s 223(7) was narrower than the scope of s 223(10) because, as Sandoz would have it, s 223(7) requires lapsing or ceasing "because of a failure to do one or more relevant acts within the time allowed" and s 223(10) requires only that there be a patent application or patent which has lapsed or ceased in relation to which a relevant act is not or cannot be done within time. The internal logic of such a scheme is not apparent to me. If (as is the case) a relevant act can be not applying for an extension of term within time and a patent can cease because its term has ended and an extension of term has not been granted at that time, then why would s 223(7), the gap creating provision, not apply to cases where the ceasing is, at one level at least, because the relevant act (applying for the extension of term within time) was not done? 147 Third, another consequence of the construction of Sandoz is that it forces a search for a gap "between the day on which the patent ceases and the day on which it is restored" as referred to in s 223(10)(b) other than the obvious gap created by s 223(7). For a patent which is extended, however, it is also the plain intention of s 79 that insofar as infringement proceedings are concerned, there is no such gap. While it may be accepted that these problems arise on Sandoz's construction only where time is extended after the term of the patent has ended, if the generic parties are right that a patent ceases within the meaning of s 223(7)-(10) when its term ends, then these problems go to the heart of the scheme as a whole. This is because s 71(2) permits an application for an extension of term at any time during the term of a patent. It must follow that, at any such time (up to the last day before the term ends), a patentee can apply for an extension of time to apply for an extension of term. While the facts of the present case may be unusual, the construction of Sandoz does not involve mere problems at the margins of the scheme. 148 For these reasons, I accept the submissions of Lundbeck, making some of the same points, as follows: 25. The argument requires the fulfilment of two conditions namely that: a. the word "ceases" in section 223(10)(b) includes the "expiry" of the original term of a patent which is extended on a date after such expiry date; and b. the word "restored" in section 223(10)(b) applies to the grant of an extension of term under section 76 of the Patents Act. 26. An immediate difficulty with the Respondents' argument is that such a construction means that section 223(10)(b) would operate to deprive the patentee of a right to sue for infringement in every case in which an extension of term is granted after expiry of the original, whether or not such extension of term involved an antecedent extension of time under section 223(2). 27. In recognition of this difficulty, the Respondents submit that section 223(10) only operates in cases where an extension of time has been sought and granted because the word "patent" in section 223(10)(b) should be construed as limited to the patent which was the subject of the "relevant act" referred to in section 223(2) and defined in section 223 that was not done within a certain time. 28. The effect of this submission is that in construing the words: "between the day on which the patent ceases and the day on which it is restored", a meaning is given to "patent" which is confined to a "patent" as referred to above via the definition of "relevant act" but there is no similar confinement of meaning in relation to the words "ceases" and "restored". 29. A consistent construction approach to all of the key words would confine "the patent ceases and the day on which it is restored" to "the patent ceases and the day on which it is restored" as referred to above. That would confine the application of section 223(10)(b) to the circumstances of section 223(7). That is a natural construction because both subsections employ the same language of "ceasing" and "restored". Moreover, the combination of ceasing and restoring is a combination used in the only other references to "restore" in the Patents Act … which relate to patent applications ceasing and the only other references to "ceasing" in the Patents Act relate to patents ceasing because of a failure to pay renewal fees. 30. In short, if it is correct to read the word "patent'' in section 223(10)(b) by reference to the internal "dictionary" of section 223, it is equally correct to read the words "ceases" and "restored" by reference to the same internal dictionary which is provided in section 223(7)…on that construction, section 223(10)(b) affords no defence to the Respondents in the present case because section 223(7) has no application to this case. … 33. Section 223(10) is a complementary provision to section 223(7). Confining the operation of section 223(10)(b) to the circumstances dealt with in section 223 is a proper approach to construction and adopting that approach, section 223(10)(b) is clearly directed at, and only at, a section 223(7) circumstance. 149 Fourth, for s 223(10) to apply it is not enough that the words "the patent ceases" be construed to apply to a patent which expires because it has reached the end of its term. It is also necessary to conclude that the patent which has expired "is restored" once an extension of term is granted. Plainly enough, the "is restored" in s 223(10) is the complement of "must be treated as having been restored" in s 223(7). No one suggests that the different wording (a requirement to treat something as having been restored and a statement that something is restored) means that s 223(10) does not operate on the circumstances created by s 223(7). One point being made in the above discussion is that Sandoz's principal construction means that it is necessary in the present case to locate a source of the 144 patent being restored outside of s 223(7). This search is by no means easy. 150 Part 3 of Ch 6 enables the term of a patent to be extended. The term may be extended before the end of the term, in which event the patent will not reach the end of its term until the end of the extended term. The term may be extended after the patent has reached the end of its term. This is a necessary consequence of s 71(2) which permits an application for an extension of term to be made at any time during the term of the patent and within the 6 months of the latest of the specified dates. As the extension of time proceedings in the present case disclose, it is also part of the statutory scheme that the second of these time requirements, but not the first, is able to be extended under s 223. But for present purposes, it is the first of these time requirements which is again significant. Section 71(2) contemplates that an application for an extension of term may be made at any time within the term of a patent. It necessarily follows that the grant of extensions of term after a patent has expired is within the contemplation of the statutory scheme. Section 79 pre-supposes this occurring. In any such case, s 79 provides that if the term expires before the extension is granted the patentee has, after the extension is granted, the same rights to start proceedings in respect of any act during the period commencing on the day of expiry and ending on the day of grant of the extension "as if the extension had been granted at the time when the act was done". 151 How is it that a patent "is restored" if, after expiry of its term, an extension of term is granted? Once a patent reaches the end of its term, the exclusive rights of the patentee under s 13 of the Patents Act come to an end. Once an extension of term is granted, the term of the patent is extended and by s 79 the patentee is vested with the same rights as if the extension had been granted at the time when the (infringing) act is done. The patent's term is extended and, as a result, s 79 creates new rights in the patentee for the intervening period. No provision of Pt 3 of Ch 6 provides for the patent to be restored either at the day of expiry or at the day of extension. To my mind, it is not a natural and ordinary reading of "is restored" in s 223(10)(b) to include within its scope a patent which expired and was later the subject of the grant of an extension of term with the consequence that s 79 applies. 152 Fifth, Sandoz relied on Beach J's description of s 223 as a "gateway" in [2017] FCA 56 at [113] so that if an extension of time is required s 223(10) operates before s 79 is engaged. According to Sandoz this conception of the scheme creates a coherent and harmonious regime in which s 79 operates subject to s 223(10). The submission is unpersuasive. That a patentee requiring an extension of time to apply for an extension of term must pass through the "gateway" of s 223(1), (2) or (2A) may be accepted. But this does not mean that s 223(10) necessarily applies. As discussed above, there is a coherent way of reading s 223(7) to 223(10) as a regime dealing with, relevantly, patents ceasing in connection with a relevant act not being done. Not every relevant act which is not or cannot be done will have a connection with a patent application lapsing or a patent ceasing. Sections 223(7) to 223(10), accordingly, are not necessarily engaged for every patent in relation to which an application for extension of time may be made under s 223(1), (2) or (2A). The fact that s 223(9) also deals with another alternative apart from a patent ceasing or a patent application lapsing, being a person exploiting an invention because of a failure to do the relevant act within the time allowed, does not undermine the coherence of the scheme I have described. 153 It is also not the case that Sandoz's characterisation of s 223 as a "gateway" which, where an extension of time is needed to apply for an extension of term, is anterior to any grant of an extension and thus before Pt 3 of Ch 6 is engaged, avoids incoherence in the statutory scheme. This is because s 79, by the fiction it creates once the extension of term is granted, effectively backdates the patentee's rights to the time when an (infringing) act is done, potentially back to the date on which the patent expired (if an infringing act is done immediately after that date). It is not a case of s 79 operating subject to s 223(10), as Sandoz would have it. Rather, one provision, s 79, gives the patentee the same rights it would have had as if the extension had been granted, potentially, on the date of expiry (if an act has been done on that date). Another provision, s 223(10), provides that infringement proceedings cannot be brought between the date on which the patent expired (that is, as Sandoz would have it, ceased) and the date on which the extension of time was granted (that is, as Sandoz would have it, the patent was restored). Ignoring the backdating effect of s 79 and saying that the extension of time must come first avoids the obvious incoherence to which Sandoz's construction gives rise. 154 For these reasons, there is no answer in the submission that "where, in order to engage s 79, the patentee has had to apply for an extension of time under s 223, and s 223(10) is engaged, as in the present case, s 79 operates subject to the effect of s 223(10) of the Act". This merely masks the reality that in such a case there would be no room for s 79 to operate at all. 155 It is also not a case of s 79 giving a patentee "superior" rights. Sandoz submitted that "s 79 cannot put the patentee in a better position with respect to statutory restrictions on its rights arising from the making of an extension of time application than a patentee who does not seek or obtain an extension of term". Nothing in my preferred construction has this effect. A patentee who does not require an extension of time to apply for an extension of term may still need the rights given by s 79 because the extension of term may not be granted before the standard term ends and the patent expires. A patentee who does require an extension of time is in no better position. The submission assumes the thing sought to be demonstrated, the application of the restriction in s 223(10). 156 Sixth, and as Lundbeck pointed out, in a scheme which permits opposition to extensions of time (s 223(6)) and to extensions of term (s 75), a consequence of Sandoz's construction is that the longer it takes a patentee to obtain an extension of time (if required) or an extension of term, then the greater the period of the gap to which s 223(10) can apply. Opponents could benefit from the s 223(10) exclusion by ensuring they oppose any extension of time and then of term, to ensure that the gap between the expiry of the patent and the patent being restored is as long as possible. A liberal construction of a remedial provision may be appropriate, but a construction which would have this consequence is not immediately attractive. 157 There is a construction available which does not give rise to any of these problems. It is to read s 223(7) to (10) as a coherent scheme, ancillary to the extension of time provisions available in s 223(1), (2) and (2A). It is to read "lapses", "ceases" and "restored" consistently throughout these provisions. It is to understand that "because of a failure to do one or more relevant acts within the time allowed" in s 223(7) requires nothing more than a connection between the relevant act not being done and the ceasing or lapsing and thus that "the patent" in s 223(10) must be the patent which has ceased or lapsed in s 223(7). Once this is accepted, the construction of Sandoz becomes untenable, for the reasons already given. This, in turn, indicates that whatever the limits of meaning of "the patent ceases" and "is restored" in s 223(10)(b), a patent expiring once its term has ended and then being extended under s 76 is not within that meaning. And once this is so, it also follows that, whatever its meaning elsewhere in the Patents Act, in the scheme of s 223(7) to 223(10), a patent does not "cease" because it expired and is not "restored" if its term is extended. 158 I am aware that this conclusion is not reconcilable with Beach J's reasoning in respect of s 223(9) in [2017] FCA 56 to the extent that Beach J concluded at [131] that the "ceasing of the patent" in s 223(9) includes the expiry of the patent. I have explained above why I am persuaded that Lundbeck's construction of s 223(10) must be preferred when the statutory scheme is considered as a whole. I also consider that once the whole of the scheme is considered it emerges that s 223(10) applies to those patent applications or patents where the gap created by s 223(7) is in play. And that gap is in play if, for example, ss 142, 143 or 143A are engaged. In any such case, there will be a lapse of a patent application or a patent will cease if the acts in those provisions are not done within the time allowed. By s 223(7), once time for the doing of those acts is extended, then the application or patent must be treated as having been restored from that time onwards. Section 223(7), unlike s 79, does not have a backdating effect. Section 223(7) creates the period on which s 223(10) can then operate. But for the reasons given s 223(7) does not apply to the 144 patent. 159 My conclusions about the preferred construction of s 223(10) also seem to me to accord with the way in which the Patents Act generally treats the expiry of a patent and the ceasing of a patent. I say generally because I accept that s 145 uses the language of "the patents … have ceased to be in force" and, in this context, this must include a patent which has expired: Regency Media Pty Ltd v MPEG LA, LLC [2014] FCAFC 183; (2014) 231 FCR 588. Statutory coherence does not require rigid consistency of use, however. Apart from s 145, when the subject is a patent, the Patents Act distinguishes between cases in which a patent has ceased and a case in which a patent's term has ended so it has expired. This is apparent from a number of examples in which there is no reason to refer to both concepts unless they have a different meaning from each other. See, in particular, ss 55(2)(b), 79(b) and (d), 79C(2)(b)(i) compared to (iii), 83(4), 85(2), 101E(1)(b), 101F(1)(c), as well as ss 143 and 143A. This difference indicates that the natural and ordinary meaning of "ceases", which may well be broad enough to encompass any form of ending of a patent, is not a secure foundation to construe ss 223(7) to 223(10). Construed in the context of the Patents Act as a whole, I am unable to accept that s 223(10)(b) applies to a patent which has expired due to its term having ended and has subsequently had its term extended. 160 It will be apparent from this discussion that the submissions of Sandoz that s 79 cannot "override" s 223(10) is also misconceived. The fact that s 223(1), (2) or (2A) applies does not necessarily mean that ss 223(7) to 223(10) applies. Section 223(10) is not overridden but simply does not apply in the present case. It follows from this that it would be wrong to conceive of s 223(10) as part of a quid pro quo for every circumstance in which s 223(1), (2) or (2A) may be available. If the concept of quid pro quo is relevant the quid and quo, in my view, are s 223(7) and s 223(10). 161 This also answers the submission of Sandoz that: The clear policy of s 223(10) (and s 223(9)) is to protect all persons adversely affected by the grant of such an extension of time. Section 223(10) should, like s 223(9), be given a liberal construction subject to that construction not being unreasonable or unnatural: see H Lundbeck A/S v Commissioner of Patents (2017) 249 FCR 41 at [100] per Beach J. 162 The problem is with the word "such" in this submission. Sandoz would have it that "such" an extension of time is an extension of time to apply for an extension of term where the patent "ceased" by reason of expiry of the term but not because of any failure to do a relevant act within the time required. But if it is fundamental to the scheme of the Patents Act that s 79 backdates the patentee's rights so that they cover the period between expiry and the grant of the extension (once the term is extended), then it is not incongruous to conclude that the protections given by s 223(10) operate on gaps within the term and have nothing to do with extensions of term. For extensions of term there is never a gap upon which s 223(10) can operate. In other words, the policy to which the Patents Act gives effect through its provisions understood as a whole is not to protect a third party from infringement proceedings relating to the period after a patent has expired and before it has been extended. The protection for third parties in the case of any "such" extension of time is s 71(2), in that irrespective of anything else an application for an extension of term "must" be made during the term of the patent, and if it is made third parties will be on notice (see s 72). It is clear from the scheme as a whole, however, that so long as this and the other conditions in s 79 are met it backdates the patentee's rights irrespective of what is said in s 223(7) and (10). 163 Accordingly, s 223(10)(b) does not apply to the 144 patent.

7.1 The issue 164 Sandoz contended in the alternative that Lundbeck AU cannot bring these proceedings because the only source of rights to do so is s 79 of the Patents Act which confers rights on the patentee alone, not an exclusive licensee of the patentee. Lundbeck contended that s 79 must be understood in the context of the patentee's rights under s 13 and the operation of s 78 and, in this context, the rights of the patentee under s 79 flow through to the exclusive licensee. 165 I am persuaded by Lundbeck's submissions. If s 79 applies, both a patentee and exclusive licensee are able to start proceedings for infringement.

7.2 Statutory provisions 166 Section 79 is in these terms: If: (a) a patentee applies for an extension of the term of a standard patent; and (b) the term of the patent expires before the application is determined; and (c) the extension is granted; the patentee has, after the extension is granted, the same rights to start proceedings in respect of the doing of an act during the period: (d) commencing on the expiration of the term of the patent; and (e) ending on the day on which the extension was granted; as if the extension had been granted at the time when the act was done. 167 It is also convenient to note here these provisions of the Patents Act: 13 (1) Subject to this Act, a patent gives the patentee the exclusive rights, during the term of the patent, to exploit the invention and to authorise another person to exploit the invention. (2) The exclusive rights are personal property and are capable of assignment and of devolution by law. 78 If the Commissioner grants an extension of the term of a standard patent, the exclusive rights of the patentee during the term of the extension are not infringed: … (b) by a person exploiting any form of the invention other than: (i) a pharmaceutical substance per se that is in substance disclosed in the complete specification of the patent and in substance falls within the scope of the claim or claims of that specification; … 120 (1) Subject to subsection (1A), infringement proceedings may be started in a prescribed court, or in another court having jurisdiction to hear and determine the matter, by the patentee or an exclusive licensee. Schedule 1 (s 3) "exclusive licensee" means a licensee under a licence granted by the patentee and conferring on the licensee, or on the licensee and persons authorised by the licensee, the right to exploit the patented invention throughout the patent area to the exclusion of the patentee and all other persons. "patentee" means the person for the time being entered in the Register as the grantee or proprietor of a patent.

7.3 Submissions 168 It is common ground that in the present case the period referred to in s 79 is the period commencing on 13 June 2009 (when the 144 patent expired) and ending on 25 June 2014 (when the extension of term of the 144 patent until 9 December 2012 was granted). Accordingly, such rights as Lundbeck may have to start proceedings for acts done between 13 June 2009 and 9 December 2012 depend on s 79. It is unnecessary to discuss the suggestion which Lundbeck might have made that the 144 patent did not expire on 13 June 2009 because, as Sandoz correctly submitted, such expiry is a pre-condition to the application of s 79. In effect, once an extension of term is granted after expiry of the term of a patent, s 79 vests the same rights in the patentee as it would have had as if the extension had been granted when the (infringing) acts were done. By this means, on the grant of an extension, the gap which would otherwise exist in the patentee's rights to sue for infringement is filled. 169 The argument for Sandoz is simple. Section 79 vests in the patentee the "same rights to start proceedings" in respect of an act during the relevant period "as if the extension had been granted at the time when the act was done". Section 79 does not vest any such rights in an exclusive licensee. Sandoz submitted that effect must be given to this clear and unambiguous language. 170 Sandoz referred to Actavis Pty Ltd v Orion Corporation [2016] FCAFC 121 in which the Full Court held that a person other than the patentee or exclusive licensee could not sue under s 120(1), saying at [251]: We think that the primary judge erred in concluding that Novartis Australia had standing to sue for infringement. In our view, s 120(1) of the Act is clear on that subject. Although s 120(1) uses the word "may", it is specific as to who may sue for infringement - the patentee or an exclusive licensee. We do not read s 120(1) as being merely indicative of the persons who can sue for infringement. The parties accept that Novartis Australia is neither the patentee nor an exclusive licensee. 171 According to Sandoz: (1) "Section 79 is, by its terms, concerned with 'rights to start proceedings.' Such rights to start proceedings for infringement are conferred separately to, and are distinct from, the 'exclusive rights' conferred by s 13 of the Act (being the rights, during the term of the patent, to exploit the invention and to authorise another person to do so)." (2) "Like s 120(1), s 79 is specific as to who has 'the same rights to start proceedings' - in this case, the patentee alone." (3) "…in this case, because the Patent expired before the extension of term was granted, an exclusive licensee does not have, after the extension is granted, the same rights (under s 120) to start proceedings in respect of the period between term expiry and the grant of the extension as if the extension had been granted at the time the act was done." (4) "The statutory purpose of the provision is found in its text, and that language must take priority over assumed, or even expressed, intent" (citing, in support, North Australian Aboriginal Justice Agency Limited v Northern Territory [2015] HCA 41; (2015) 256 CLR 569 at [229] and SZOFE v Minister for Immigration and Citizenship [2010] FCAFC 79; (2010) 185 FCR 129 at [62]). (5) "In any event, there is logical sense in the proposition that only the patentee may sue during the period provided for in s 79(d)-(e) because during that period the 'patentee' has no patent in force and therefore no exclusive rights under s 13 over which it could grant an exclusive licence. If the extension of term is granted before the end of the extended term then the patentee will again have patent rights which could be the subject of an exclusive licence." 172 Lundbeck submitted that: (1) "…section 79 is a substantive provision, which has the effect that the patentee's rights under the patent backdate to the date of expiry. The consequence is that the exclusive rights of the patentee including the right to sue for infringement extend to acts of infringement during the period after the expiry of the original term of the patent and before the extension is granted. In circumstances where there is an exclusive licensee which has the benefit of an exclusive licence of the patentee's rights, this extension of the duration of the patentee's rights flows on, via the exclusive licence, to the exclusive licensee." (2) Section 78 supports this construction. On Lundbeck's construction, s 78, in referring to the exclusive rights of the patentee, applies to the patentee and the exclusive licensee. The rights of both are confined by s 78. On the approach of the generic parties the limitations imposed by s 78 would not apply to exclusive licensees. The approach gives different meanings to the term "patentee" in ss 78 and 79 and thus should be rejected.

7.4 Discussion 173 Section 79 vests in the patentee the "same rights to start proceedings" the patentee would have had as if the extension had been granted at the time of doing the (infringing) act. Those rights are the rights under s 120. 174 If, as in the present case, the alleged infringing acts commenced on the date the 144 patent expired, s 79 operates to vest in Lundbeck A/S (the patentee) from 25 June 2014 (the date the extension was granted) the same rights it has to start proceedings under s 120 as it would have had to start proceedings in respect of the alleged infringing acts as if the extension had been granted on 13 June 2009. The proceedings are thus started under s 120, not under s 79. 175 It follows that the argument of Sandoz is that although the rights in s 120(1) are vested in the patentee and the exclusive licensee, because the patentee's rights during the extended term depend on s 79 which refers only to the patentee having the "same rights" during the period between the expiry of the patent and the grant of the extension of term, it is the patentee alone who has the rights under s 120 as the patentee would have had if the extension had been granted at the relevant earlier time. This is important because, through s 79, s 120 is in play. 176 Lundbeck's argument based on s 78 assumes that the sections are dealing with the same rights. This is not immediately apparent. Section 78 is dealing with the "exclusive rights of the patentee" which, but for s 78, may be infringed. The only rights these can be are the rights to exploit the invention under s 13. Section 79 refers to the rights to start proceedings for infringement. But as explained below, unless the s 13 rights exist, there can be no infringement. These apparent complexities are sufficient to undermine Sandoz's proposition that s 79 is clear and unambiguous. To the contrary, as explained below, s 79 is meaningless unless the right to "start proceedings" is or carries with it the exclusive s 13 rights to exploit the invention. 177 In common with all statutory provisions, s 79 cannot be construed in isolation. It must be understood in the context of the statutory scheme as a whole. The statutory scheme includes ss 13, 78, 79 and 120. These provisions are to be construed appreciating the operation of s 12 of the Acts Interpretation Act 1901 (Cth) (that every section of an Act shall have effect as a substantive enactment) and with regard to s 15AA of that Act (in interpreting a provision of an Act, the interpretation that would best achieve the purpose or object of the Act is to be preferred to each other interpretation). Section 15AB of that Act, concerning extrinsic material, is also relevant in that under s 15AB, consideration may be given to extrinsic material to assist in ascertaining the meaning of a provision if the provision is "ambiguous or obscure" (s 15AB(1)(b)(i)) or "the ordinary meaning conveyed by the text of the provision taking into account its context in the Act and the purpose or object underlying the Act leads to a result that is manifestly absurd or is unreasonable" (s 15AB(1)(b)(ii)). 178 It follows that it would be wrong to construe ss 13, 78, 79 and 120 as if they take effect sequentially merely because, in the case of a particular patent, extension of a term of a patent must occur after the initial grant. All of these provisions are part of the Patents Act. All operate in relation to a patent immediately on its grant. The fact that ss 78 and 79 deal with future contingencies does not mean that the sections may be ignored when attempting to understand the rights the legislation gives to patentees (and thus which may be assigned to exclusive licensees). For example, ss 13 and 78 deal with the same rights, the exclusive rights of the patentee to exploit the invention. Accordingly, it is plainly necessary to read ss 13 and 78 together. 179 By s 13 the patentee has the exclusive rights to exploit the invention during the term of the patent. These rights are assignable including by way of an assignment which creates an exclusive licensee. It goes without saying that the patentee cannot assign any greater right than that which it has vested in it by the Patents Act which includes not just s 13 but also s 78. By s 78 the exclusive rights of the patentee during an extension of term are not infringed by a person exploiting the invention other than in the circumstances identified. Section 78 qualifies the (future and contingent) rights of the patentee from the moment of grant of the patent. The provisions are not to be read as if s 78 operates only if and when an extension of term is granted. 180 It necessarily follows from this that if the rights of the patentee are not infringed because of s 78, the rights of the exclusive licensee are also not infringed because of s 78. Section 78 assumes, as is the case, that a patentee can never assign to an exclusive licensee any rights greater than the patentee has. 181 Accordingly, Lundbeck's approach to s 78 seems to assume that the rights a patentee has under s 13 before an extension are not subject to s 78 from the moment of grant of the patent. It is this assumption which permits Lundbeck to argue that, if "patentee" does not include exclusive licensee in s 78 (and s 79), s 78 will confine the rights of the patentee during an extension of term but not of the exclusive licensee. 182 In my view, the true position is that from the moment a patent is granted ss 13 and 78 (and, as discussed below, ss 79 and 120) all have effect in relation to the patent. As such, a patentee's rights are always subject to the contingent operation of s 78 and no greater right may be assigned to an exclusive licensee at any point in time. And as a result of this, s 78 operates to confine the rights of the patentee and, of necessity, the exclusive licensee if and when an extension of term is granted. 183 Other complexities are apparent in the relationship between the provisions. In particular: (1) Section 78 assumes the patentee's exclusive s 13 rights to exploit the invention exist. Section 78 does not itself grant any such rights. It merely confines the s 13 rights of the patentee which it assumes otherwise exist by operation of ss 13, 79 and 120. This reinforces the point made above that s 78 has effect on the s 13 rights that can be assigned on and from the grant of a patent. (2) Section 79 also assumes the patentee's exclusive s 13 rights to exploit the invention during the period with which it deals (between expiry of the patent and grant of the extension) exist. This must be so or s 79 cannot function. Section 79 gives the patentee the "same rights" to start proceedings as it would have had in respect of an act during the relevant period as if the extension had been granted at the time of the act. The rights the patentee would have had to start proceedings on this hypothesis are the exclusive rights to exploit the patent and the right to start proceedings for infringement under s 120. This reinforces the necessity of recognising that s 79 too has effect in relation to a patent immediately on the grant of the patent, in the sense that s 79 applies to every patent, albeit that the rights it creates are future contingent rights. (3) While the language of s 78 is dealing with the period "during the term of the extension", s 78 also assumes the existence of s 79 and assumes that s 79 operates to give the patentee the exclusive s 13 rights to exploit the invention during the period with which s 79 deals (between expiry of the patent and the grant of the extension). Section 78 must so assume or otherwise the patentee would have the right to start proceedings under s 79 (for infringement of its exclusive right to exploit the invention, which s 79 assumes exist) and s 78 would not apply to that period (between expiry of the patent and the grant of the extension) to confine the patentee's rights. Putting it another way, s 78 functions coherently only if the right of the patentee to start proceedings given under s 79 is understood as meaning or carrying with it the patentee's exclusive rights under s 13 to exploit the invention. Otherwise, the patentee would be vested by s 79 with the "same right" to "start proceedings" but would not have the foundational exclusive rights under s 13 to exploit the invention to enable any such proceedings to be started. Without the foundational exclusive rights under s 13 existing during the period between expiry of the patent and the grant of the extension, s 79 is meaningless for the patentee (let alone the exclusive licensee). 184 The source of the patentee's s 13 rights during the term of an extension is clear. However, the source of the patentee's s 13 rights during the period between expiry of the patent and the grant of the extension is unclear, because the language of s 79 relates to the right to "start proceedings". But as I have tried to explain, this right is meaningless for the patentee unless it carries with it the exclusive rights to exploit the invention during the period covered by s 79. 185 To reprise, when the term of the patent is extended under s 76, s 13 gives the patentee the exclusive rights during the extension, confined by s 78. This is because during the extended term is still "during the term of the patent" as specified in s 13. It follows that if the extension is granted before the patent expires, there will be no period which is not "during the term of the patent" as provided for in s 13. The "term" of the patent will continue seamlessly from the ordinary term into the extended term. Subject to s 78, the patentee and exclusive licensee may start proceedings for infringement under s 120 during the entirety of the term of the patent, both ordinary and extended. 186 If, however, the extension is granted after the patent expires, then the position is different. There are two possibilities in this regard. The extension may be granted before the last day of the extended term or after the last day of the extended term. If the extension is granted before the last day of the extended term, then there is no issue for the period of the extended term after the grant of the extension. That period will be "during the term of the patent" and thus the patentee (and thus the exclusive licensee) has the ss 13, 78 and 120 rights. But in such a case there is a gap between the date of expiry of the patent and the grant of the extension. And an extension may also be granted after the last day of the extended term in which event there will also be a gap (and there will be no period during the term of the patent after the grant of the extension). Section 79 is concerned with the gap which will exist in both cases. 187 Section 79 deals with the gap by creating a legal fiction. The fiction involves treating the extension as granted at the time the (infringing) acts were done when in fact, the acts must have occurred within the gap between expiry of the patent and the grant of the extension of the term. The rights which s 79 vests in the patentee under the legal fiction are expressed to be the same rights to start proceedings the patentee would have had if the extension had been granted when the act was done. As explained, for the fiction to work for the patentee, those rights (to start proceedings), to be the "same rights" the patentee would have had, necessarily carry with them the patentee's exclusive s 13 rights to exploit the invention. That is, the s 79 fiction can only function at all if part of the fiction is that during the period covered by s 79, when the extension of term is granted, the patentee has the s 13 exclusive rights to exploit the invention as if it were "during the term of the patent", when in fact that period is never "during the term of the patent". And because, like ss 13 and 78, s 79 has effect (albeit a future contingent effect) from the moment a patent is granted, it must be the case that the suite of rights which a patentee has from the date of grant of a patent are the exclusive rights provided for by ss 13, 78 and 79 (and s 120). To operate as a coherent scheme, it is that suite of rights which a patentee may assign to an exclusive licensee, including the exclusive rights as qualified by s 78 and as provided for by the legal fiction in s 79. 188 These considerations mean that s 79 is an ambiguous provision for at least two reasons. One, it does not refer to the patentee having the same exclusive rights to exploit the invention as if the extension had been granted at the time the act was done, even though this must be the case. Two, it refers instead to the patentee having the "same rights" to start proceedings as it would have had if the extension had been granted at the time the act was done when, on my view as to how the section must be understood (as giving the patentee the same right to exploit the invention), s 120 would simply operate according to its terms. 189 The one thing that is clear is that, given the ambiguity of s 79, the interpretation which best gives effect to the purpose or objects of the Patents Act is to be preferred and regard may be had to extrinsic material to assist in the task of giving meaning to the provision. In particular, this is not a case where, like Actavis v Orion, the statutory provision is clear and, in terms, does not permit a person other than a patentee or the exclusive licensee to start proceedings. 190 The parties did not identify any extrinsic material which is of much assistance. Clause 77 of the Explanatory Memorandum, Patents Bill 1990 (Cth) states this: Clause 77: Rights of patentee 109. This clause has the effect that, where a patent's term is extended after the original term expired, the patentee's rights will backdate to the date of expiry. This provision recognises that interested third parties will have been warned, before the term expired, of the pending application for an extension, by way of the notice referred to in subclause 70(3). 191 To my mind, the objects or purposes of the Patents Act are best given effect by reading s 79 as if it vests in the patentee the same exclusive rights it would have had to exploit the invention under s 13 during the period covered by s 79 as if the extension had been granted at the time when the (infringing) acts were done. On this basis, on an assignment of the s 13 rights to an exclusive licensee, the rights assigned are subject to and include the future contingent effects of ss 78 and 79. The words "the patentee has … the same rights to start proceedings …" in s 79 would not be read as if they operated to exclude an exclusive licensee from starting proceedings. They would be read as rights the patentee has from the grant of the patent (as with ss 13, 78 and 120) which are capable of assignment to an exclusive licensee so that for the patentee in fact to have the "same rights to start proceedings", both the patentee and the exclusive licensee must have the rights under s 120. To read the words as if they were intended to exclude an exclusive licensee from starting proceedings under s 120 for the period covered by s 79 is to ensure that the patentee does not have the "same rights" for which s 79 provides because, under the fiction created by s 79, those rights include assignment and, if assigned to an exclusive licensee, the rights under s 120. 192 Sandoz's arguments to the contrary would result in a scheme which is unworkable. Section 79 cannot be read literally as to do so means the patentee has no exclusive rights to protect by starting proceedings for infringement in the period which s 79 covers. So the "same rights to start proceedings" cannot be given its literal meaning. It must mean (also or instead) that the patentee has the same exclusive rights under s 13 it would have had if the extension had been granted at the earlier time. Not only would s 79 not work if s 79 is given its literal meaning, s 78 also would not work because the period covered by s 79 is not "during the term" of the patent and thus the confining effect of s 78 would not operate during the period covered by s 79. So, even for the patentee and leaving aside any consideration of an exclusive licensee, s 79 cannot be given its literal meaning. For these reasons, contrary to the submissions for Sandoz, there is no "logical sense in the proposition that only the patentee may sue during the period provided for in s 79(d)-(e) because during that period the 'patentee' has no patent in force and therefore no exclusive rights under s 13 over which it could grant an exclusive licence". 193 The best way to give effect to the scheme of the Patents Act for extensions of term of patents is to construe ss 78 and 79 in the same manner. For the reasons given above, the rights of the patentee which are confined under s 78 necessarily, from the grant of the patent, are also so confined so that the exclusive licensee can never get a greater right and takes rights subject to s 78. And the rights of the patentee under s 79, in my view, must be understood as contingent rights that existed from the moment of grant of the patent, so that the exclusive licensee takes those rights as well. Otherwise, looked at from the time of grant of the patent when the entire suite of rights exists and is assignable, the patentee would not have the "same rights to start proceedings" as it would have had as provided for in s 79. To be the "same rights to start proceedings" the patentee must be able to create an exclusive licensee who can also start proceedings under s 120 including for the period of an extension. The object or purpose of s 79 being to fill the gap between the expiry of the patent and the grant of the extension by giving rights as if the extension had been granted at the earlier time is best achieved by a construction which results in the same position for the patentee during the s 79 period. The construction of Sandoz does not achieve that object or purpose. 194 For these reasons I accept Lundbeck's submission that s 79 is a substantive provision which has the effect that the patentee's rights under the patent backdate to the date of expiry and, where there is an exclusive licensee which has the benefit of an exclusive licence of the patentee's rights, this extension of the duration of the patentee's rights flows on, via the exclusive licence, to the exclusive licensee. 195 Accordingly, my view is that Lundbeck AU, as well as Lundbeck A/S, was entitled to start proceedings for acts done between 13 June 2009 and 9 December 2012 on and from the grant of the extension on 25 June 2014.

8.1 The issue 196 Sandoz contended, in the further alternative, that Lundbeck AU was not the exclusive licensee of the 144 patent at any material time. Lundbeck contended that the agreement between Lundbeck A/S and Lundbeck AU dated 26 September 2005 constituted Lundbeck AU as the exclusive licensee of the 144 patent, and no subsequent event varied Lundbeck AU's status as such. 197 I consider Lundbeck AU to be the exclusive licensee of the 144 patent.

8.2 Generic parties' submissions 198 Sandoz put the argument this way: There are three related bases on which the Respondents allege that Lundbeck AU was not the exclusive licensee during the extended term: (a) First, by reason of the settlement agreement made between Lundbeck DN and Lundbeck AU and Sandoz Pty Ltd in or about February 2007 (CB Vol H4, Tab 268, p 4667) and/or the letter dated 28 May 2009 from Corrs to Jim Sharkey of Sandoz (CB Vol H4, Tab 298, p 4583). As to this matter, the Respondents propose to adopt and rely on Sandoz's closing submissions, to be supplemented as necessary by oral submissions. Importantly … the Sandoz Licence was granted jointly and severally by Lundbeck DN and Lundbeck AU. That is, Lundbeck AU (the purported exclusive licensee) and Lundbeck DN (the parent company of Lundbeck AU) each independently granted Sandoz a licence to exploit the Patent in an agreement that each was a party to and concurred in. (b) Secondly, it is clear that the "Licence Agreement" relied on by Lundbeck and entered into by Lundbeck DN and Lundbeck AU on 26 September 2005 (CB Vol G, Tab 93, p 1884) does not fully disclose the legal rights and entitlements between them, and does not constitute the whole of the arrangement between them. When the entirety of the relationships among Lundbeck DN, Lundbeck AU, and third parties authorised by one or both of them to exploit the Patent is considered, it becomes apparent that Lundbeck AU was not the exclusive licensee during the extended term, in accordance with the definition of "exclusive licensee" in Schedule 1 to the Act. Those relationships demonstrate the creation or recognition, by Lundbeck DN and Lundbeck AU, of rights to exploit the Patent in Australia that are in legal effect and practical operation inconsistent with the rights that must be held by a person in order to be an exclusive licensee under the Act. (c) Thirdly, and relatedly to (a) and (b), the Court ought to conclude that the effect of clause 6 of the Licence Agreement (which prohibits Lundbeck AU from sub-licensing third parties) is to deny Lundbeck AU exclusive licensee status in the circumstances. The Respondents do not submit that a purported exclusive licensee must have the right to sub-licence, in order to meet the requirements of the Act. However, in a case where (i) the purported exclusive licensee is contractually prohibited from authorising a third party to exploit the invention and (ii) the patentee and exclusive licensee have jointly and severally purported to authorise a third party to exploit the invention, it necessarily follows that those parties are jointly conducting themselves contrary to the terms of the purported licence, have not excluded the patentee from exercising the exclusive right to authorise another to exploit the invention, and further or alternatively have varied the agreement between themselves to permit the patentee to do so: see Bristol-Myers Squibb Company v Apotex Pty Ltd (2015) 228 FCR 1 at [103].

8.3 Discussion 199 It is not in dispute that in order to be the exclusive licensee the licence must confer on the exclusive licensee all of the rights to exploit the patent in the patent area to the exclusion of all other persons including the patentee and that these rights include the right to authorise another person to exploit the invention: Bristol-Myers Squibb Co v Apotex Pty Ltd [2015] FCAFC 2; (2015) 228 FCR 1 at [103]-[105]. 200 It is also the case, as Sandoz submitted, that there may be cases in which the manner in which the exclusive licensee may exercise its plenary rights is regulated by agreement and cases in which a purported plenary right of an exclusive licensee is not, taking into account all relevant provisions of applicable agreements, not such a plenary right at all (see, for example, Actavis v Orion at [234]). In the latter case, the necessary conclusion is that the purported exclusive licensee is not the exclusive licensee. As Sandoz also noted, in Actavis v Orion the unilateral character of the subsequent inconsistent rights granted by the patentee was crucial. Thus, at [232], the Full Court said: The existence of what might be conflicting rights conferred by Orion [the patentee] on another in respect of the supply of a generic product in an agreement post-dating the 2014 licence cannot change the effect of that licence. There has been no variation of the 2014 licence; and there is no reason to give the 2014 licence a scope that is more limited than its language bears, because of later conduct of one party to it involving a third party. 201 It is also not in dispute that, if the 26 September 2005 agreement is considered in isolation, it does (Lundbeck's case) or purports to (Sandoz's case) vest in Lundbeck AU all of the rights to exploit the 144 patent in the patent area to the exclusion of all including the patentee. In that agreement Lundbeck A/S as licensor, by cl 2, grants to Lundbeck AU as licensee "the right to Exploit the invention the subject of the Patent throughout the Patent Area to the exclusion of the Licensor and all other persons, subject to the provisions of this Deed". In the agreement Exploit and Patent Area are defined to have the meaning given in Sch 1 to the Patents Act, and Patent is defined as the 144 patent. By cl 3 the licence continues until termination under cl 4(a), which provides that Lundbeck A/S may terminate the Deed for any reason on giving 30 days' notice in writing. By cl 6, Lundbeck AU may not assign or sub-licence the rights under the agreement to a third party. 202 I accept Lundbeck's submission that the 26 September 2005 agreement, on its face, constitutes Lundbeck AU as the exclusive licensee of the 144 patent. To resolve the arguments of Sandoz that the 2005 agreement does not reflect all legal rights between Lundbeck A/S and Lundbeck AU or that those rights were subsequently varied, it is necessary to consider the other agreements on which Sandoz relied to support its three propositions as recorded above. As Sandoz noted, Lundbeck A/S is the parent company. Lundbeck AU is a wholly owned subsidiary of Lundbeck A/S. There are other members of the Lundbeck group including CNS Pharma which is a wholly owned subsidiary of Lundbeck AU. 203 In February 2007 Lundbeck A/S (defined as Lundbeck Denmark), Lundbeck AU and Sandoz entered into an agreement (referred to as the Sandoz settlement agreement). To the extent currently relevant, the provisions of the Sandoz settlement agreement include recitals as follows: (a) Lundbeck Denmark is the owner of the Patent. (b) Lundbeck Australia is the exclusive licensee of the Patent in Australia. (c) On 13 April 2006, Sandoz commenced the Proceeding. (d) Without Sandoz conceding the validity of any claims in the Patent or Lundbeck Denmark or Lundbeck Australia conceding the invalidity of any claims in the Patent, the Parties have agreed to settle the Proceedings on the terms set out in this Agreement. 204 The defined terms include Patent (the 144 patent) and Proceeding (NSD 714 of 2006 by Sandoz against Lundbeck Denmark seeking revocation of the Patent). 205 By cl 3(1) of the Sandoz settlement agreement: Lundbeck Denmark and Lundbeck Australia jointly and severally grant Sandoz an irrevocable non-exclusive license to the Patent effective from: (a) 31 May 2009 if the Patent expires on 13 June 2009; (b) 26 November 2012 if the Patent expires on 9 December 2012; (c) 31 May 2014 if the Patent expires on 13 June 2014; (d) 2 weeks prior to the expiry of the Patent if the Patent expires on a date other than a date described in clause 3(a) to (c). 206 By cl 3(2) of the Sandoz settlement agreement: In addition to the licence granted under clause 3(1), Lundbeck Denmark and Lundbeck Australia jointly and severally grant Sandoz an irrevocable non-exclusive licence to the Patent, effective from the beginning of the calendar month in which the licence granted under clause 3(1) becomes effective, for the sole purpose of manufacturing, importing, marketing and offering to sell (but not selling or supplying) pharmaceutical products containing escitalopram. 207 On 28 May 2009 the solicitors for Lundbeck A/S and Lundbeck AU sent a letter which Sandoz contends varied the Sandoz settlement agreement. That letter is considered below. The Sandoz settlement agreement either does or does not have the effect for which Sandoz contended, which is to disclose that Lundbeck AU was not the exclusive licensee of the 144 patent by reason of the 2005 agreement at any time or ceased to be the exclusive licensee of the 144 patent by reason of the Sandoz settlement agreement. The former argument depends on the pleaded proposition that the 2005 agreement does not fully disclose the legal rights between Lundbeck A/S and Lundbeck AU and does not constitute the whole of the arrangement between them (in contrast to Actavis v Orion at [226]-[229]). The latter argument depends on Lundbeck A/S and Lundbeck AU having jointly and severally agreed in the Sandoz settlement agreement to grant Sandoz a licence of the 144 patent which must mean that rights to do so were agreed between Lundbeck A/S and Lundbeck AU to remain in Lundbeck A/S, a necessary corollary of which is that Lundbeck AU ceased to be the exclusive licensee from that time. This is said to be supported by the fact that by cl 6 of the 2005 agreement Lundbeck AU was prohibited from authorising a third party to exploit the 144 patent, yet Lundbeck AU did so under the Sandoz settlement agreement. 208 The same considerations apply to the other agreements and documents on which Sandoz relied. The documentary evidence is summarised as follows by Sandoz: (a) As at January 2005, Lundbeck AU was the exclusive distributor in Australia of Lundbeck DN's CIPRAMIL and LEXAPRO products: GX TB, Ex 32, Tab 1. (b) Under cl 6 of the "Licence Agreement" dated 26 September 2005, which is said to constitute the exclusive licence, Lundbeck AU is prohibited from authorising a third party to exploit the Patent: CB Vol G, Tab 93, p 1887. (c) In the "Sandoz Settlement Agreement" executed by the parties in February 2007, Lundbeck DN and Lundbeck AU jointly and severally granted Sandoz a licence to exploit the Patent: CB Vol H4, Tab 268, p 4669. (d) Lundbeck AU has sold Lundbeck's LEXAPRO products in Australia since at least June 2009: CB Vol A, Tab 5, p 62 [7(c)]. (e) CNS Pharma has supplied Lundbeck's ESIPRAM products in Australia since at least June 2009: see GX TB, Ex 32, Tabs 21, 31 and 33 and CB Vol B, Tab 64, para 17; T571.40-41 and LCS [2]. (f) In "confirmatory" documents formalising the existing supply arrangements between Lundbeck DN and Lundbeck AU, and Lundbeck DN and CNS Pharma, Lundbeck AU and CNS Pharma are treated equally, vis-à-vis Lundbeck DN, in respect of escitalopram products: GX TB, Ex 32, Tabs 37 and 38. (g) In each of those confirmatory documents, Lundbeck DN purports to grant authority to the counterparty under "the Patents": GX TB, Ex 32, Tabs 37 and 38, cl 2.1 in each agreement. In each of these agreements, "Patents" is defined to include "any present…patent…relating to the Finished Products that are Lundbeck owned or licensed or otherwise covered by a right of use of Lundbeck". There is no evidence of any grant of a right of use by Lundbeck AU to Lundbeck DN … (h) In Lundbeck's internal documents, and documents prepared for the purposes of tax authorities, Lundbeck Australia is treated as a mere sales and distribution company and is described as "not holding any IP": see GX TB, Ex 32, Tabs 39 (sections 1.1, 2.2, 2.3, 5), 40 (sections 2.2.2.1, 2.8.1.3, 2.8.2.2, 2.8.4, and especially 2.10) and 41 (section 1.1.3). The 2005 "Exclusive Licence" agreement is not mentioned in the "overview of inter-company agreements" in the Lundbeck AU Transfer Pricing Files: see GX TB, Ex 32, Tab 39, "Section 5". (i) In June 2009, Lundbeck Export A/S, of which Lundbeck DN is the parent company, licensed Genepharm (Australia) Limited to purchase via CNS Pharma, and sell in Australia, escitalopram products. The recitals to that agreement indicate that Lundbeck DN has granted Lundbeck Export A/S the rights to commercialise the product in Australia. Lundbeck Export A/S executed that agreement on 12 June 2009: GX TB, Ex 32, Tab 20. 209 According to Sandoz this evidence supports the following: (a) Lundbeck DN and Lundbeck AU, in entering into the Sandoz Settlement Agreement, necessarily agreed between themselves, or acted pursuant to an arrangement or understanding between them, to deal with the Patent inconsistently with the terms of the Licence Agreement in at least two ways: (i) Lundbeck DN's authority to severally grant a licence to Sandoz to exploit the Patent is inconsistent with Lundbeck AU being an exclusive licensee under the Licence Agreement. (ii) Lundbeck AU's authority to severally grant a licence to Sandoz to exploit the Patent is inconsistent with the prohibition in clause 6 of the Licence Agreement on sub-licensing a third party. (b) It follows from (a) that the Licence Agreement does not represent the entirety of the relationship between Lundbeck DN and Lundbeck AU and is not an instrument that records or discloses the full legal rights and entitlements of the parties to it. (c) As submitted, there is no evidence of any grant by Lundbeck AU to Lundbeck DN of a licence to exploit the Patent. Further, because Lundbeck AU was not itself granted the right to authorise third parties to exploit the Patent, it could not grant Lundbeck DN the right to authorise third parties to do so (even if it could have licensed Lundbeck DN to exploit the Patent). Accordingly, any grant by Lundbeck DN to a third party of authority to exploit the Patent can only be pursuant to retained rights to authorise third parties to do so, or an understanding between Lundbeck DN and Lundbeck AU that Lundbeck DN was permitted to do so. (d) The conduct of Lundbeck DN and AU in February 2007 in entering into the Sandoz Settlement Agreement is sufficient to undermine Lundbeck AU's status as exclusive licensee. They jointly assented, in a written agreement, to the grant by each of them jointly and severally of authority to exploit the Patent, and in doing so necessarily departed from the Licence Agreement for the reasons set out in (a)-(c) above. (e) In the documents formalising Lundbeck DN's pre-existing arrangements vis-à-vis Lundbeck AU and CNS Pharma in respect of, inter alia, escitalopram products in 2012, Lundbeck DN is treated as the holder of the relevant patents and Lundbeck DN purports to authorise Lundbeck AU and CNS Pharma respectively to exploit those patents. By the equal treatment of Lundbeck AU and CNS Pharma, Lundbeck DN is necessarily dealing with each of them as no more than a non-exclusive licensee. By definition, it cannot be dealing with both as exclusive licensees. (f) The proposition that Lundbeck AU is a mere seller and distributor of escitalopram products is supported by Lundbeck's internal documentation and documentation prepared for the purpose of satisfying taxation authorities about the group's internal arrangements. (g) That proposition is also supported by the "Genepharm Supply Agreement" …As recorded in the recitals to that agreement, Lundbeck DN granted Lundbeck Export A/S rights to commercialise an escitalopram product in Australia. Lundbeck Export A/S executed the document on 12 June 2009. As within the Lundbeck group, Lundbeck DN purported, on or before 12 June 2009, to grant Lundbeck Export A/S the right to deal in an escitalopram product in Australia. That would amount to a licence from the patentee to exploit the Patent and undermine Lundbeck AU's status as exclusive licensee. It is also further conduct inconsistent with Lundbeck AU's alleged status as exclusive licensee within the group. 210 Sandoz also submitted this: To the extent that it is necessary to go any further, those matters … support a finding that Lundbeck DN and Lundbeck AU intended that the Licence Agreement would give Lundbeck AU standing to sue under the Act, but did not intend that Lundbeck AU would in fact be the exclusive licensee of the Patent. That is, although they intended the Licence Agreement to give rise to one legal result, they did not intend to deal with the Patent consistently with the underlying division of rights that was necessary to give rise to that legal result. 211 I do not accept Sandoz's case. 212 In Actavis v Orion at [226]-[229] the issue was the lack of any pleading that the agreement between the patentee and exclusive licensee was a sham. As [229] of Actavis v Orion notes, a sham transaction is one involving a subterfuge. In the present case, Sandoz has avoided the language of sham, but claims that the 2005 agreement does not fully disclose the legal rights between Lundbeck A/S and Lundbeck AU and does not constitute the whole of the arrangement between them. There are only two ways of understanding this case. Either the 2005 agreement does not fully disclose the legal rights between Lundbeck A/S and Lundbeck AU at the time it was made, in which event the 2005 agreement was a sham or it should be inferred that Lundbeck A/S and Lundbeck AU subsequently varied the 2005 agreement to create different rights from those created by the 2005 agreement. It appears from the last of the submissions recorded above, (that by the 2005 agreement Lundbeck A/S and Lundbeck AU did not intend that Lundbeck AU in fact would be the exclusive licensee of the 144 patent) that Sandoz's claim is unnecessarily coy; it is contending that the 2005 agreement is a sham transaction. 213 To the extent the case of Sandoz is that the 2005 agreement is a sham, certain principles are engaged which Sandoz has not confronted. In particular, as Leeming JA (McColl JA and Sackville AJA agreeing) said in Lewis v Condon; Condon v Lewis [2013] NSWCA 204; (2013) 85 NSWLR 99: [60] Basic to the legal notion of sham is that it is a confined and exceptional aspect of the process of giving legal meaning to a document, as Professor Conaglen has pointed out ("Sham Trusts" (2008) 67 CLJ 176 at 206): The relevance of the sham doctrine, and the difference between it and normal processes of construction, lies in the fact that it justifies the court in ignoring (as opposed to construing) the usual primary material regarding that transaction, and focusing its attention instead on all other material factors which indicate the arrangement that the parties in fact intended. [61] That echoes the words of Windeyer J in Scott v Cmr of Taxation (Cth) (No 2) (1966) 40 ALJR 265 at 279: The difficult and debatable philosophic questions of the meaning and relationship of reality, substance and form are for the purposes of our law generally resolved by asking did the parties who entered into the ostensible transaction mean it to be in truth their transaction, or did they mean it to be, and in fact use it as, merely a disguise, a facade, a sham, a false front … concealing their real transaction. [62] The sham doctrine is thus one of those relatively rare doctrines in the law where legal meaning is given to a document by reference to a subjective intention. Other examples are a plea of non est factum at law and a claim for rectification in equity. All these doctrines "must necessarily be kept within narrow limits", for all subtract from the objective theory of contractual obligation, and if unchecked would cause "serious mischief": see Toll (FGCT) Pty Ltd v Alphapharm Pty Ltd [2004] HCA 52; (2004) 219 CLR 165 at [46]-[47]. This has long been the law: see for example Jordan CJ's reasons in Perpetual Trustee Co (Ltd) v Bligh (1940) 38 SR NSW 33 at 39-40. In all these areas, strong evidence is required in order to displace the orthodox approach to construction. Hence the "heavy onus" that must be discharged by the plaintiff in a non est factum case (Petelin v Cullen (1975) 132 CLR 355 at 360) and the need for "clear and convincing proof" in a rectification suit (Franklins Pty Ltd v Metcash Pty Ltd [2009] NSWCA 407; (2009) 76 NSWLR 603 at [451]-[460]). [63] Because a finding of sham requires a finding of an intent to deceive, considerations associated with Briginshaw v Briginshaw (1938) 60 CLR 336 require a cautious approach: Raftland Pty Ltd v Commission of Taxation [Raftland Pty Ltd v Commissioner of Taxation [2008] HCA 21; (2008) 238 CLR 516] at [36]. Thus there is a "strong and natural presumption against holding a provision or a document a sham": National Westminster Bank plc v Jones [2001] 1 BCLC 98 at [59] (Neuberger J). "A court will only look behind a transaction's ostensible validity if there is a good reason to do so, and 'good reason' is a high threshold, since a premium is placed on commercial certainty": Official Assignee v Wilson [2007] NZCA 122; [2008] 3 NZLR 45 at [52] (Robertson and O'Regan JJ). Lockhart J referred to "a strong finding, and one which cannot be made if another inference is at least equally open" in Sharrment Pty Ltd v Official Trustee in Bankruptcy at 461. 214 Sandoz has not come close to proving that the 2005 agreement was a sham when made. Its case does not grapple with the relevant principles which apply. Avoiding the language of sham, when the substance of the case is that there was a sham, does not mean that the applicable principles can also be avoided. There is no foundation for any finding that, when the 2005 agreement was made, Lundbeck A/S and Lundbeck AU, "did not intend that Lundbeck AU would in fact be the exclusive licensee" of the 144 patent. Nor is there any foundation for a finding that, at that time, Lundbeck A/S and Lundbeck AU did not intend to deal with the 144 patent consistently with the 144 patent. 215 The alternative case, as noted, is that subsequent events prove or give effect to some other agreement, arrangement or understanding between Lundbeck A/S and Lundbeck AU. 216 All of the other agreements are to be construed in accordance with the following principles, summarised in Franklins Pty Ltd v Metcash Trading Ltd [2009] NSWCA 407; (2009) 76 NSWLR 603 as follows: [14] The state of the law in this respect is to be ascertained from a number of High Court cases: Maggbury Pty Ltd v Hafele Australia Pty Ltd (2001) 210 CLR 181; 185 ALR 152; 53 IPR 1; [2001] HCA 70 at [11]…; Pacific Carriers [Pacific Carriers Ltd v BNP Paribas [2004] HCA 35; (2004) 218 CLR 451] at [22]; Zhu v Treasurer (NSW) (2004) 218 CLR 530; 211 ALR 159; [2004] HCA 56 at [82]…; Toll [(FGCT) Pty Ltd v Alphapharm Pty Ltd [2004] HCA 52; (2004) 219 CLR 165] at [40] and International Air Transport Association v Ansett Australia Holdings Ltd (2008) 234 CLR 151; 242 ALR 47; 65 ACSR 1; [2008] HCA 3 at [8] and [53] … These cases are clear. The construction and interpretation of written contracts is to be undertaken by an examination of the text of the document in the context of the surrounding circumstances known to the parties, including the purpose and object of the transaction and by assessing how a reasonable person would have understood the language in that context. There is no place in that structure, so expressed, for a requirement to discern textual, or any other, ambiguity in the words of the document before any resort can be made to such evidence of surrounding circumstances. … [19] The essential character of the task of construction of commercial contracts can be seen in a number of authoritative decisions of the High Court, and of other courts authoritatively endorsed by the High Court. A commercial contract should be given a businesslike interpretation … Thus, the nature and extent of the commercial aims and purposes of the agreement or parts thereof are part of the essential background circumstances … The need for a businesslike construction not only informs the nature and extent of the extrinsic material legitimately of assistance, but it also directs the approach to be taken to the ascription of meaning to the words used by the parties. The words should be given a construction so as "to avoid … [making] commercial nonsense or is shown to be commercially inconvenient": Hide & Skin Trading Pty Ltd v Oceanic Meat Traders Ltd (1990) 20 NSWLR 310 at 313-14 per Kirby P cited by the court in Zhu at [82] . This is not only a reflection of the place of the informing surrounding circumstances, it is also a requirement not to approach words in a business contract pedantically or in a manner prone to defeat the evident commercial purpose. 217 It is true that as at January 2005 Lundbeck A/S and Lundbeck AU had entered into a distribution agreement. The distribution agreement is dated 1997 (referred to as the 1997 agreement). Under it, Lundbeck A/S granted to Lundbeck AU the right to distribute, promote, market and sell Finished Products in the Territory (art 2). The Finished Products are defined to mean the products of Lundbeck origin listed in Appendix A "supplied by Lundbeck in final consumer packages ready for resale and use in the Territory". The Finished Products in Appendix A include but are not limited to Lexapro and Cipramil. The Territory is Australia. 218 It is apparent that the 1997 agreement does not deal with patent rights. If I infer that there is no other such agreement (which I do) then I would construe the 1997 agreement as conferring such rights on Lundbeck AU as may be necessary for it to do that which the 1997 agreement authorises. As nothing in the 1997 agreement excludes any rights of Lundbeck A/S, the 1997 agreement cannot constitute Lundbeck AU as the exclusive licensee of the 144 patent. Nor does Lundbeck suggest this to be so. As a result, all that can be said is that Lundbeck was not the exclusive licensee of the 144 patent until the 2005 agreement was executed. There is no inconsistency between the 1997 agreement and the 2005 agreement making Lundbeck AU the exclusive patentee of the 144 patent. The 2005 agreement does not commit Lundbeck A/S to supply or Lundbeck AU to distribute any product. It is the 1997 agreement which, construed as a whole (see, for example, the definition of Finished Product and arts 6 and 7) involves obligations of supply and distribution. 219 Nothing about the 1997 agreement suggests that the 2005 agreement is a sham or is to be read as if it said something other than what it says. 220 The Sandoz settlement agreement was entered into in February 2007, after execution of the 2005 agreement. Accordingly, the 2005 agreement is part of the surrounding circumstances in which the 2007 agreement must be construed. Further, it is not apparent to me why an agreement entered into two years after the 2005 agreement would be taken to be evidence that the 2005 agreement, at the time it was entered into, was a sham or did not mean what it said. As a result, Sandoz is left with an argument that the Sandoz settlement agreement, by its terms, evidences some variation to the 2005 agreement pursuant to which Lundbeck A/S and Lundbeck AU agreed that Lundbeck A/S had the right to authorise Sandoz to exploit the 144 patent (so Lundbeck AU was no longer the exclusive licensee of the 144 patent) and Lundbeck AU also had the right to authorise Sandoz to exploit the 144 patent (despite cl 6 of the 2005 agreement). 221 The problem with this argument is that there are good reasons not to infer any such agreed variation between Lundbeck A/S and Lundbeck AU by reason of the terms of the Sandoz settlement agreement or otherwise. 222 First, the Sandoz settlement agreement, in recital (b), states that "Lundbeck Australia is the exclusive licensee of the Patent in Australia". There is no recital suggesting that by entering into the Sandoz settlement agreement or otherwise Lundbeck A/S and Lundbeck AU had agreed that Lundbeck AU was no longer to be the exclusive licensee of the 144 patent. Accordingly, the case of Sandoz is that, from the terms of the Sandoz settlement agreement (specifically, the joint and several grant of the licence in cl 3(1)), this should be inferred to be the mutual intention of Lundbeck A/S and Lundbeck AU. 223 Second, such an inferred mutual intention from the joint and several grant of the licence to Sandoz seems objectively inconsistent with how a reasonable person would have understood the language of cl 3(1) in the surrounding circumstances which were known to Lundbeck A/S, Lundbeck AU and Sandoz. The known surrounding circumstances are referred to in cl 1 of the Sandoz settlement agreement in which Sandoz agreed not to assist Alphapharm in proceedings NSD 1120 of 2005 or NSD 1870 of 2005 or Arrow in proceeding NSD 954 of 2006. These are the proceedings which Lindgren J decided in the Lindgren J judgment. In those proceedings, Lundbeck A/S as patentee and Lundbeck AU as exclusive licensee cross-claimed against Alphapharm for infringement. These circumstances, referred to by Lindgren J in the Lindgren J judgment at [34] and [634], were part of the known surrounding circumstances to the parties at the time they entered into the Sandoz settlement agreement. 224 Third, and related to the second proposition, imputing to Lundbeck A/S and Lundbeck AU (and, for that matter, Sandoz) a mutual intention, by the terms of the Sandoz settlement agreement, to vary the 2005 agreement so that Lundbeck AU was no longer the exclusive licensee makes no commercial sense. From the perspective of Lundbeck A/S and Lundbeck AU they would be agreeing to change the status of Lundbeck AU so that Lundbeck AU would no longer be able to maintain the infringement proceedings against Alphapharm, which are the very proceedings which Sandoz, in the Sandoz settlement agreement, was agreeing not to provide any assistance. From the perspective of Sandoz, Lundbeck AU would no longer be the exclusive licensee contrary to recital (b) of the Sandoz settlement agreement. Why a reasonable person would infer from the Sandoz settlement agreement any variation to the 2005 agreement in these circumstances is not apparent. 225 Fourth, there are ways of construing the Sandoz settlement agreement which do not make it internally inconsistent (as between recital (b) and cll 3(1) and (2)) and do not involve apparent commercial unreasonableness given the known surrounding circumstances. 226 One such construction is to read cll 3(1) and 3(2) as a form of "belt and braces" drafting by Sandoz (it being known that an in-house lawyer for Sandoz, James Sharkey, drafted the settlement agreement). That is, the clauses are to be read as saying, in effect, that to the extent either or both of Lundbeck A/S and Lundbeck AU are capable of granting a licence to Sandoz, each does so individually and together. As the true legal position is that in the 2005 agreement, the fact is Lundbeck A/S had no such capacity. To that extent, cll 3(1) and 3(2) are simply inoperative. Only Lundbeck AU could grant the licence and thus all that Lundbeck A/S could (and did) agree to was to endorse Lundbeck AU to do so notwithstanding cl 6 of the 2005 agreement. This construction makes commercial sense from the perspective of all of the parties to the Sandoz settlement agreement. Sandoz had no interest in the arrangements between Lundbeck A/S and Lundbeck AU other than it knew that Lundbeck A/S and Lundbeck AU considered Lundbeck AU to be the exclusive licensee of the 144 patent (see recital (b)). Sandoz's interest was in obtaining a legally effective licence from whichever Lundbeck party could grant it. From Sandoz's perspective a joint and several grant from both covered all possibilities. From Lundbeck's perspective, the same result could be achieved without undermining the recited status of Lundbeck AU as the exclusive licensee. 227 Another such construction also reflects this "belt and braces" approach by Sandoz to which Lundbeck could agree without undermining the recited status of Lundbeck AU as the exclusive licensee. Again, it is based on the self-evident proposition that Sandoz had no interest in the arrangements between Lundbeck A/S and Lundbeck AU. If to this is added another surrounding circumstance, that Sandoz had no control over arrangements between Lundbeck A/S and Lundbeck AU after entry into the Sandoz settlement agreement, then obtaining a joint and several licence protected Sandoz from any future termination of the agreement between Lundbeck A/S and Lundbeck AU. 228 On this basis, cll 3(1) and 3(2) should be construed as meaning not that Lundbeck A/S has purported to grant a licence which it had no power to grant, but that the grant by Lundbeck A/S is a future contingent grant, the relevant contingency being that Lundbeck AU no longer has exclusive or any power to grant Sandoz the licence. From Sandoz's perspective, ensuring its rights did not depend on arrangements between Lundbeck A/S and Lundbeck AU, was a commercial necessity. From the perspective of Lundbeck A/S and Lundbeck AU, cll 3(1) and (2) also ensured that the Sandoz licence was not dependent on the existing arrangements between Lundbeck A/S and Lundbeck AU continuing. 229 It is not necessary to determine which of these constructions of the Sandoz settlement agreement is to be preferred. The relevant point is that both are to be preferred to the position posited by Sandoz, that the Sandoz settlement agreement had the effect of making Lundbeck AU no longer the exclusive licensee of the 144 patent or otherwise evidences another agreement between Lundbeck A/S and Lundbeck AU under which Lundbeck AU was no longer the exclusive licensee of the 144 patent. The fact that it may be taken by the Sandoz settlement agreement that Lundbeck A/S and Lundbeck AU must have agreed that Lundbeck AU could sub-license rights despite cl 6 of the 2005 agreement (noting cl 9 of the 2005 agreement which provides for any variation to be in writing and signed by the parties) is immaterial. The Sandoz settlement agreement is in writing and signed by Lundbeck A/S and Lundbeck AU. The capacity for Lundbeck AU to enter into the Sandoz settlement agreement was a necessity for the settlement agreement to be made. But to construe the Sandoz settlement agreement, by cll 3(1) and 3(2), making or meaning that Lundbeck AU is no longer the exclusive licensee is of a different character in that it: (1) is not necessary to give a sensible commercial effect to the Sandoz settlement agreement; (2) would create an internal inconsistency between recital (b) and cll 3(1) and 3(2) of the Sandoz settlement agreement; and (3) would produce a commercially unreasonable outcome from the perspective of all parties to the Sandoz settlement agreement having regard to recital (b). 230 Just as there is no reason to construe the Sandoz settlement agreement in this way, so too there is even less reason for inferring from the Sandoz settlement agreement that Lundbeck A/S and Lundbeck AU, at any time had or entered into some other agreement, arrangement or understanding pursuant to which Lundbeck AU was not or was no longer the exclusive licensee of the 144 patent. As between Lundbeck A/S and Lundbeck AU such an agreement, arrangement or understanding at any time between the date of the 2005 agreement and the date of the Sandoz settlement agreement would have been commercially unfathomable given that Lundbeck AU being the exclusive licensee was the source of its right to maintain the infringement proceedings against Alphapharm which remained undetermined. 231 The same reasoning applies to the 28 May 2009 letter. That letter involves undertakings from Sandoz to Lundbeck A/S and Lundbeck AU and cross-undertakings from Lundbeck A/S to Sandoz including that "Lundbeck has also agreed that, pending the outcome of the Full Court appeal, Sandoz may import into Australia pharmaceutical products containing escitalopram". The reference to "Lundbeck" here is a reference to Lundbeck A/S. It would not be inferred from this that Lundbeck AU was not the exclusive licensee of the 144 patent, for the reasons already given. This is particularly so given that the reality was that Sandoz had already imported the products into Australia purportedly under the licence it had in cl 3(2) of the Sandoz settlement agreement. Lundbeck A/S was not granting Sandoz any rights in its cross-undertaking. It was confining what Sandoz had already done. As such, the 28 May 2009 letter does not add anything to the Sandoz settlement agreement, insofar as the issue of the status of Lundbeck AU is concerned. 232 The supply agreement dated 22 June 2009 is between a third party, Genepharm (Australia) Pty Ltd and Lundbeck Export A/S. The recitals to that agreement record that Lundbeck A/S is the parent company of Lundbeck Export A/S. They record also that Lundbeck's affiliated companies have been marketing escitalopram in Australia under the trademark Lexapro since 2003. The recitals also record that: … Lundbeck's parent company H. Lundbeck A/S … has granted Lundbeck [Lundbeck Export A/S] rights to commercialise the said product in i.a. Australia [sic]. …Genepharm is interested in purchasing finished, packaged pharmaceutical products from Lundbeck, via its affiliate CNS Pharma, in order for Genepharm to sell the products in Australia under the trademark Esipram®… 233 Other relevant provisions of the 22 June 2009 agreement include: 2.1 Lundbeck hereby grants to Genepharm and Genepharm hereby accepts the exclusive right to import, market, distribute and sell the Product in the Territory under the Trademark at its own risk and account. … 2.2 Lundbeck may perform its obligations under this agreement through its affiliate, CNS Pharma. 234 By cl 1, CNS Pharma is the same company which is a party in this matter, Product means escitalopram oxalate tablets, Territory means Australia, and Trademark means Esipram. 235 Other facts forming part of the surrounding circumstances to the 22 June 2009 agreement are relevant. In short, CNS Pharma was (and is) the wholly owned subsidiary of Lundbeck AU which was selling in Australia escitalopram oxalate tablets under the name Esipram. Leaving aside two critical facts, one that Lundbeck A/S is not a party to the 22 June 2009 agreement and two that the agreement does not mention rights under the 144 patent (cl 2.1 confining itself to "right[s] … under the Trademark") there are two possibilities which appear to inform the case of Sandoz, being: (1) despite Lundbeck Export A/S not being the patentee or the exclusive licensee of the 144 patent, it was purporting by the 22 June 2009 agreement to grant rights under the 144 patent which it did not have to Genepharm; and (2) as a result of (1), it must or should be inferred that Lundbeck A/S and Lundbeck AU had varied the 2005 agreement so that Lundbeck AU was no longer the exclusive licensee of the 144 patent, as Lundbeck Export A/S was also a licensee of that patent. 236 Again, both possibilities appear to me to involve commercial unrealities. It is not necessary or commercially sensible to read the recital that Lundbeck A/S has granted Lundbeck Export A/S the right to "commercialise the said product in… Australia" as meaning that Lundbeck A/S and Lundbeck AU had varied the 2005 agreement so that Lundbeck AU was no longer the exclusive licensee of the 144 patent. Lundbeck A/S and Lundbeck AU are not parties to the 22 June 2009 agreement and it cannot be read as if they were. Nor, for the same reason, should it be read as evidencing some varied agreement, arrangement or understanding between Lundbeck A/S and Lundbeck AU as to Lundbeck AU's status as the exclusive licensee of the 144 patent, particularly not to detriment of those parties where other commercially sensible inferences are available. 237 In particular, Lundbeck AU must be taken to have authorised its wholly owned subsidiary, CNS Pharma, to exploit the invention the subject of the 144 patent. CNS Pharma did so under the trademark Esipram. The notion that another company in the Lundbeck group, Lundbeck Export A/S, could purport to authorise Genepharm to sell Esipram in Australia under the Esipram trademark, including agreeing with Genepharm that it, Lundbeck Export A/S, could perform its obligations under the agreement though its affiliate, CNS Pharma, without Lundbeck AU also agreeing to licence Genepharm to do so is a commercial and legal nonsense. This may explain why no rights under the 144 patent are purported to be granted to Genepharm under cl 2.1. Only Lundbeck AU could grant those rights given its status as the exclusive licensee of the 144 patent. Alternatively, if cl 2.1 is read as Lundbeck Export A/S giving Genepharm rights under the 144 patent, then it must be inferred that Lundbeck AU, in its capacity as the exclusive licensee, gave such rights to Lundbeck Export A/S. There is no other way Lundbeck Export A/S could have obtained such rights unless it is inferred that, despite being in the midst of the appeal from the Alphapharm proceedings, Lundbeck A/S and Lundbeck AU agreed that Lundbeck AU should no longer be the exclusive licensee of the 144 patent. Again, to my mind, any such inference would be commercially unfathomable. 238 Whatever the actual position of Lundbeck AU, the inescapable facts are these. Neither Lundbeck A/S nor Lundbeck AU are parties to the 22 June 2009 agreement. The mere fact that Lundbeck Export A/S, from the terms of the recitals may be taken to be a subsidiary of Lundbeck A/S, does not mean that Lundbeck Export A/S had the power unilaterally to vary the 2005 agreement between Lundbeck A/S and Lundbeck AU. On the same basis, the statement in the recital that Lundbeck A/S had granted Lundbeck Export A/S the right to commercialise escitalopram oxalate tablets in Australia cannot evidence, as between Lundbeck A/S and Lundbeck AU, any variation to the 2005 agreement given that neither Lundbeck A/S nor Lundbeck AU is a party to the 22 June 2009 agreement. It follows that if cl 2.1 involves any right to exploit the invention the subject of the 144 patent, unless Lundbeck Export A/S was on a frolic of its own (which also seems unlikely), it must be taken to have had Lundbeck AU's permission, to the extent necessary, to authorise Genepharm in accordance with cl 2.1 of the 22 June 2009 agreement. 239 On this basis, at worst, an agreement between a company related to Lundbeck A/S and Lundbeck AU has entered into an agreement with a third party which may contain an error in the first recital insofar as it refers to Lundbeck A/S and not Lundbeck AU having granted Lundbeck Export A/S rights to commercialise the product in Australia. That error is not attributable to Lundbeck A/S or Lundbeck AU. It cannot found inferences of the kind which Sandoz asked to be drawn of either the 2005 agreement being a sham or having been subsequently varied so that Lundbeck AU is no longer the exclusive licensee. 240 The fact that CNS Pharma, a wholly owned subsidiary of Lundbeck AU, has supplied Esipram tablets in Australia since at least 2009 is immaterial to the status of Lundbeck AU as the exclusive licensee of the 144 patent. As noted, the only available inference is that, consistent with its rights as the exclusive licensee of the 144 patent, Lundbeck AU has authorised its wholly owned subsidiary to exploit the invention. 241 As I would understand it, Sandoz stressed cl 6 of the 2005 agreement (that Lundbeck AU may not assign or sub-license its rights to a third party) because Lundbeck AU subsequently did so, in the case of Sandoz and CNS Pharma, and (arguably) Lundbeck Export A/S (as discussed above). I have already noted above that I see no difficulty in construing the Sandoz settlement agreement as involving a variation to cl 6 of the 2005 agreement as between Lundbeck A/S and Lundbeck AU, and that I do not see acceptance of this variation as providing any support for a further variation by which Lundbeck AU was agreed between them no longer to be the exclusive licensee of the 144 patent. The same reasoning applies to the inferred agreement between Lundbeck AU and CNS Pharma and (if it is necessary to say it exists) between Lundbeck AU, Lundbeck Export A/S and Genepharm. And insofar as cl 9 of the 2005 agreement is relevant (variations were to be in writing and signed by the parties), Lundbeck A/S and Lundbeck AU were free to vary that requirement as well. Given the relationships between members of the Lundbeck group, an inference of variations to cll 6 and 9 to the extent necessary is far more supportable than the inference Sandoz proposed that Lundbeck A/S and Lundbeck AU had an agreement, arrangement or understanding either that the 2005 agreement was a sham or that it was varied so Lundbeck AU was no longer the exclusive licensee of the 144 patent. 242 The so-called "confirmatory" documents formalising the existing supply agreements with Lundbeck A/S entered into in 2012 do not assist Sandoz. First, they are supply and distribution agreements under which Lundbeck A/S is to supply and either Lundbeck AU or CNS Pharma is to distribute the products. Second, they are entered into in circumstances where Lundbeck AU, the parent of CNS Pharma, is constituted as the exclusive licensee of the 144 patent under the 2005 agreement. Third, they say nothing about the 2005 agreement or Lundbeck AU's status as the exclusive licensee of the 144 patent under the 2005 agreement. Fourth, they are not confined to escitalopram products. They cover numerous products of which escitalopram products are but one and, in that sense, are to be understood as pro-forma documents in respect of products where rights under relevant Patents and Trademarks (as defined) may be different as between products. 243 In accordance with my reasoning above, I would not infer from the distribution agreement between Lundbeck A/S and CNS Pharma, entered into when Lundbeck A/S was seeking to extend the term of the 144 patent, that Lundbeck A/S and Lundbeck AU should be inferred to have some agreement, arrangement or understanding by which Lundbeck AU was no longer the exclusive licensee of the 144 patent. As noted, CNS Pharma was the wholly owned subsidiary of Lundbeck AU. As between Lundbeck A/S and Lundbeck AU it makes far more commercial sense to construe cl 2.1 of the agreement with CNS Pharma in one of two ways rather than taking it as evidence of some agreement, arrangement or understanding by which Lundbeck AU was no longer the exclusive licensee of the 144 patent. One, to the extent the 144 patent is concerned, Lundbeck A/S had no rights under the 144 patent to grant to CNS Pharma. Two, to the extent it had any such rights, Lundbeck A/S got those rights from Lundbeck AU as the exclusive licensee of the 144 patent. 244 In short, I do not accept the case theory of Sandoz that Lundbeck A/S and Lundbeck AU should be found to have created a sham in 2005. As such, Lundbeck AU was the exclusive licensee of the 144 patent in 2005. I also do not accept the case theory of Sandoz that Lundbeck A/S and Lundbeck AU should be found subsequently to have acted contrary to their own commercial interests by varying the 2005 agreement so that Lundbeck AU was no longer the exclusive licensee when, at all material times, Lundbeck A/S and Lundbeck AU were positively asserting to the contrary in the Alphapharm proceedings or were taking active steps to extend the term of the 144 patent for the sole possible reason of seeking damages for infringement which, for Lundbeck AU, depended on its continuing status as the exclusive licensee of the 144 patent. I accept that they must be taken to have varied the 2005 agreement to permit Lundbeck AU to authorise others (CNS Pharma and perhaps Genepharm, if the agreement with Genepharm is read in that way) to exploit the invention the subject of the 144 patent. But, the latter conclusion is essential and, in all of the surrounding circumstances as described and as known to those involved, makes commercial sense. The former conclusion is one of a number of alternatives which would have made no commercial sense in the circumstances as they existed at all relevant times. As a result, Sandoz's construction of the various agreements is inconsistent with the required approach, which is to read the language of those agreements as would be read by a reasonable person, in the context of the known surrounding circumstances, and so as to achieve commercial sense rather than commercial nonsense. 245 The other documents on which Sandoz relied, in my view, are immaterial. Whatever might have been said in taxation or internal documents, it is the agreements which found the rights. Further, many of these documents, although undated, must constitute subsequent conduct. The transfer pricing documents relate to the 2008-2010 period, 2011 and 2015. As such, nothing said in them can affect the construction of the various agreements (made between 2005 and 2012) made previously. The fact that Lundbeck AU is described as a "distributor" is also immaterial. The context of the transfer pricing documents is different from that of patent rights. In particular, nothing Lundbeck AU said in a transfer pricing document for 2015 can be relevant. The fact that the 2015 document says that Lundbeck A/S "owns" all IP and that Lundbeck AU "owns no IP", even if it could be relevant, is consistent with Lundbeck A/S being the patentee of the 144 patent and Lundbeck AU being the exclusive licensee of that patent. An analysis of the roles of Lundbeck A/S and Lundbeck AU cannot bear upon any matter relevant for present purposes. Nor can it be relevant that these documents do not mention the 2005 agreement between Lundbeck A/S and Lundbeck AU. 246 For these reasons I do not accept the case of Sandoz that Lundbeck AU was not the exclusive licensee of the 144 patent at any material time. It was the exclusive licensee of the 144 patent under the 2005 agreement at all material times.

9.1 The issues 247 Sandoz contended that: (1) by cll 3(1) and (2) of the Sandoz settlement agreement Lundbeck A/S and Lundbeck AU granted Sandoz a licence to exploit the invention the subject of the 144 patent which, either because of Lindgren J's decision on 24 April 2008 in [2008] FCA 559 (Sandoz's principal construction) or because the 144 patent expired on 13 June 2009 (Sandoz's alternative construction), commenced on 31 May 2009 (the general licence to exploit under cl 3(1)) and 1 May 2009 (the confined licence to manufacture, import, market, and offer to sell only under cl 3(2) respectively); (2) on 28 May 2009 this licence was varied by cross-undertakings under which Sandoz obtained a licence even if the 144 patent expired on 9 December 2012 and under which Lundbeck A/S is liable to Sandoz for damages; (3) the Sandoz settlement agreement contains a release of Sandoz which Lundbeck has breached in various ways including by bringing these proceedings against Sandoz; and 248 Lundbeck contended that: (1) the general licence it granted to Sandoz under cl 3(1) of the Sandoz settlement agreement commenced on 26 November 2012, being two weeks before the date when the 144 patent ultimately expired in accordance with the extension of term granted on 25 June 2014 to 9 December 2012. Accordingly, the confined licence under cl 3(2) commenced on 1 November 2012. Alternatively, if the licence commenced on 31 May 2009 it also expired on 13 June 2009 when, on the case of the generic parties, the patent expired; (2) the licence was not varied on 28 May 2009; (3) it has not breached the release in the Sandoz settlement agreement. 249 I have concluded that the licence "to the Patent" commenced on 31 May 2009 but also expired on 13 June 2009, with the expiry of the patent.

9.2 The terms of the Sandoz settlement agreement 250 It is convenient to identify the relevant terms of the Sandoz settlement agreement, despite some of those terms being set out above. 251 The terms include the following: Recitals (a) Lundbeck Denmark is the owner of the Patent. (b) Lundbeck Australia is the exclusive licensee of the Patent in Australia. (c) On 13 April 2006, Sandoz commenced the Proceeding. (d) Without Sandoz conceding the validity of any claims in the Patent or Lundbeck Denmark or Lundbeck Australia conceding the invalidity of any claims in the Patent, the Parties have agreed to settle the Proceeding on the terms set out in this Agreement. Definitions In this Agreement, each of the following terms has the meaning ascribed: Agreement means this settlement agreement. Parties means the parties to the Agreement. Patent means Australian Letters Patent Number 623144. Proceeding means Federal Court of Australia proceedings number NSD 714 of 2006 commenced by Sandoz in the New South Wales District Registry of the Federal Court of Australia against Lundbeck Denmark seeking revocation of the Patent. Operative Provisions 1 Release by, and obligations of, Sandoz Sandoz: (a) agrees to not provide assistance, nor to assist or procure other persons to provide assistance, to: (i) Alphapharm Pty Limited in Federal Court of Australia Proceeding number NSD 1120 of 2005, or Federal Court of Australian Proceedings number 1870 of 2005; or (ii) Arrow Pharmaceuticals Limited in Federal Court of Australia Proceedings number 954 of 2006; (b) agrees to not commence an Australian Court proceeding seeking revocation of any of the claims of the Patent, as they exist at the date of this Agreement; (c) agrees to keep confidential the terms of this Agreement indefinitely and to keep confidential the existence of this Agreement until 20 February 2007. To avoid doubt, as from 20 February 2007, the mere fact that the parties have agreed to settle the Proceeding is not confidential; (d) releases Lundbeck Denmark from all claims, actions and causes of actions (including any claim for costs), present and future, relating to the Proceeding; (e) agrees that Lundbeck Denmark may plead this Agreement to bar any claim, action or cause of action (including any claim for costs) brought by Sandoz relating to the Proceeding; (f) agrees to not commence or maintain any claim, action or cause of action (including any claim for costs) against Lundbeck Denmark relating to the Proceeding (with the exception of a claim, action or cause of action for breach or enforcement of this Agreement); and (g) agrees to indemnify Lundbeck Denmark against any liability, loss or costs arising from a breach of clause 1(f) above. 2 Release by, and obligations of, Lundbeck Lundbeck Denmark and Lundbeck Australia jointly and severally: (a) agree to keep confidential the terms of this Agreement indefinitely and to keep confidential the existence of this Agreement until 20 February 2007. To avoid doubt, as from 20 February 2007, the mere fact that the parties have agreed to settle the Proceeding is not confidential. (b) release Sandoz from all claims, actions and causes of actions (including any claim for costs), present and future, relating to the Proceeding; (c) agree that Sandoz may plead this Agreement to bar any claim, action or cause of action (including any claim for costs) brought by Lundbeck Denmark or Lundbeck Australia relating to the Proceeding; (d) agree to not commence or maintain any claim, action or cause of action (including any claim for costs) against Sandoz relating to the Proceeding (with the exception of a claim, action or cause of action for breach or enforcement of this Agreement); and (e) agree to indemnify Sandoz against any liability, loss or costs arising from a breach of clause 2(d) above. 3 Licence to exploit the Patent (1) Lundbeck Denmark and Lundbeck Australia jointly and severally grant Sandoz an irrevocable non-exclusive licence to the Patent effective from: (a) 31 May 2009 if the Patent expires on 13 June 2009; (b) 26 November 2012 if the Patent expires on 9 December 2012; (c) 31 May 2014 if the Patent expires on 13 June 2014; or (d) 2 weeks prior to the expiry of the Patent if the Patent expires on a date other than a date described in clause 3(a) to (c). (2) In addition to the licence granted under clause 3(1), Lundbeck Denmark and Lundbeck Australia jointly and severally grant Sandoz an irrevocable non-exclusive licence to the Patent, effective from the beginning of the calendar month in which the licence granted under clause 3(1) becomes effective, for the sole purpose of manufacturing, importing, marketing and offering to sell (but not selling or supplying) pharmaceutical products containing escitalopram. (3) For the avoidance of doubt, nothing in this Agreement is to be taken as granting a licence of, or authorisation to exploit, any patent other than the Patent. … 5 Finalisation of the Proceeding (a) By 20 February 2007, the Parties will take all steps necessary to finalise the Proceeding on the basis that the Proceeding is to be discontinued with no order for costs. (b) For the purpose of clause 7(a), but without limiting the Parties' obligations under that clause: (i) Lundbeck Denmark will arrange for the Notice of Discontinuance which is Annexure A to this Agreement to be signed and returned to the solicitors for Sandoz, Shelston IP Lawyers; and (ii) Sandoz will file and serve that Notice of Discontinuance on 20 February 2007. … 7 Miscellaneous (a) This Agreement constitutes the entire agreement of the Parties about its subject matter and supersedes all previous agreements, understandings and negotiations on that subject matter. … (c) Each Party acknowledges it enters into this Agreement voluntarily and after obtaining legal advice about the content and effect of this Agreement. …

9.3 The competing constructions 252 Clause 3(1): the key words are "if the Patent expires on" in cl 3(1)(a) to (d) of the Sandoz settlement agreement. Three alternative constructions of the words "if the Patent expires on" in cl 3(1)(a) to (d) are proposed: (1) Sandoz's principal construction: the words mean, in effect, "if it is determined in the Alphapharm proceedings (including any appeal in those proceedings) that the 144 patent expires on one or other of the nominated dates"; (2) Sandoz's alternative construction (supported by Apotex and Aspen): the words mean what they say so that, if at any time the 144 patent expires, the licence in cl 3(1) commences (and thus the licence in cl 3(2) also commences); or (3) Lundbeck's construction: the words mean the ultimate expiry of the 144 patent including as a result of any extension of term of the patent, whether or not the extension is granted before or after the end of the standard term of 20 years. 253 I should say immediately that to focus on the words "if the Patent expires on" in cl 3(1)(a) to (d) without also recognising that the licence granted is a licence "to the Patent", as stated in the preamble to the clause, would be inappropriate. Further, the licence is not a licence to exploit the invention the subject of the patent. 254 Clauses 2(b) and (c): the key words in dispute are "relating to the Proceeding". Sandoz contended that the "Proceeding" as referred to in cl 2 (being Sandoz's proceeding to revoke the 144 patent), is broad enough to encompass any claim by Lundbeck that Sandoz infringed the 144 patent, and thus cl 2 operates as a bar to Lundbeck bringing the present infringement proceedings. Lundbeck contended that that the scope of the "Proceeding" was confined to Sandoz's claim to revoke the 144 patent, so that the releases in cll 2(b) and (c) cannot extend beyond that subject-matter.

9.4 Circumstances surrounding the Sandoz settlement agreement 255 Both Sandoz and Lundbeck identified that there were surrounding circumstances relevant to the construction of the Sandoz settlement agreement. This approach reflects the fact that contracts are to be construed objectively, their meaning being that which would be conveyed by their terms to the reasonable person, to whom is attributed the relevant background knowledge which "would reasonably have been available to the parties in the situation in which they were at" at the time of the Sandoz settlement agreement: Maggbury Pty Ltd v Hafele Australia Pty Ltd [2001] HCA 70; (2001) 210 CLR 181 at [11]. 256 A further aspect of the principles of contractual construction should be noted. The point is exposed in the following extract from Franklins Pty Ltd v Metcash Trading Ltd [2009] NSWCA 407; (2009) 76 NSWLR 603 at [24] per Allsop P (as he was): [24] The High Court authorities …in particular Pacific Carriers v BNP Paribas and Toll, and the recognition of the significance of the objective theory assist in appreciating the scope of the evidence that is admissible. The evidence, to be admissible, must be relevant to a fact in issue, probative of the surrounding circumstances known to the parties or of the purpose or object of the transaction, including its genesis, background, context and market in which the parties are operating. What is impermissible is evidence, whether of negotiations, drafts or otherwise, which is probative of, or led so as to understand, the actual intentions of the parties. Such evidence might be legitimate, however, if directed to one of the legitimate aspects of surrounding circumstances. The distinction can be subtle in any particular case. As Macfarlan JA and I said in Kimberley Securities Ltd v Esber [2008] NSWCA 301 at [5]: The possible subtlety of the distinction can be seen in Lord Wilberforce's reasons in Prenn v Simmonds [Prenn v Simmonds [1971] 1 WLR 1381] … at 1384-1485, and the recognition that the objective commercial aim may, possibly, be ascertained from some aspect of what has passed between the parties. The distinction can also be seen in what Mason J said in Codelfa [Codelfa Construction Pty Ltd v State Rail Authority (NSW) (1982) 149 CLR 337] at 352 about prior negotiations and their legitimate use "to establish objective background facts which were known to both parties and the subject matter of the contract", and their inadmissibility "in so far as they consist of statements and actions of the parties which are reflective of their actual intentions or expectations" … 257 Accordingly, evidence of pre-contractual negotiations may prove relevant surrounding circumstances but is not otherwise a legitimate aid to construction. In particular, to the extent that such evidence discloses only the state of mind (be it in the form of subjective knowledge, understanding or intention) of an individual involved in the negotiation or drafting of the Sandoz settlement agreement, such evidence is immaterial. In this regard, I also note cl 7(a) of the Sandoz settlement agreement which says that it "constitutes the entire agreement of the Parties about its subject matter and supersedes all previous agreements, understandings and negotiations on that subject matter". 258 Apart from this, it was common ground that subsequent conduct cannot be used as an aid to construction: see, for example, Metcash at [6]-[11]. Accordingly, to the extent that Sandoz referred to it having acted on the basis that it had a licence under cl 3(1) of the Sandoz settlement agreement from May 2009, that evidence is also irrelevant. Similarly, to the extent that Lundbeck referred to Sandoz having acted inconsistently with it having a licence under cl 3(1) of the Sandoz settlement agreement from May 2009, that evidence is irrelevant. 259 I accept that the relevant circumstances include the various steps in the Alphapharm proceedings to the extent that those matters existed before on or about 16 February 2007 when the Sandoz settlement agreement was executed. 260 I also consider that the provisions of the Patents Act must form part of the relevant background knowledge of the parties because knowledge of those provisions was reasonably available to the parties before they entered into the agreement. Whatever one or other person subjectively thought about these provisions and their potential application to the present case is immaterial for the purpose of contractual construction. The provisions existed. Accordingly, that Lundbeck A/S might try and might be able to avail itself of the provisions to apply for an extension of time to apply for an extension of term of the 144 patent based on the registration of Cipramil was at least an objective possibility. 261 For what it is worth, it is also apparent that this possibility was actually known to Lundbeck A/S as a result of advice it received from Watermark patent attorneys on 14 July 2005. I do not suggest this can be relevant to the task of construction but it supports my conclusion that this possibility was knowledge reasonably available to the parties. 262 This said, I accept that the knowledge reasonably available to the parties as at the time of the Sandoz settlement agreement was not knowledge, for example, that: section 223(2) would apply to the circumstances; Lundbeck A/S would apply for and be granted an extension of time to apply for an extension of term of the 144 patent based on the Cipramil registration; or Lundbeck A/S would be granted an extension of term of the 144 patent based on the Cipramil registration. 263 None of these things were known. They were mere possibilities. In particular, the time for applying for an extension of term based on the registration of Cipramil had long since passed. Cipramil had been registered on 9 December 1997. The period of six months from that date as referred to in s 71(2) of the Patents Act ended on 9 June 1998, nearly a decade before the Sandoz settlement agreement. Lundbeck's application for an extension of term based on the registration of Lexapro was made on 22 December 2003, so even that application was well out of time if the relevant date under s 71(2) was the registration of Cipramil. It also would not have been obvious from the combined operation of the provisions of the Patents Act and Regulations that the period of six months in s 71(2) was capable of extension. Although the High Court ultimately found this period to be capable of extension in [2014] HCA 42 (by a majority of 3:2), it could not be suggested that the parties would have shared any such understanding at the time of the Sandoz settlement agreement. Further, it would not have been obvious that Lundbeck A/S could bring itself within the scope of the extension of time provision in s 223(2) which depends on an error or omission by the person concerned. And even it if managed to overcome all of these hurdles, Lundbeck A/S would still have to persuade the Commissioner to exercise a discretion in its favour that it be granted a very lengthy extension of time when Lundbeck A/S had known of Alphapharm's contention that the registration of Cipramil was the relevant date under s 71(2) since at least 6 July 2005 (when Alphapharm commenced proceedings) and knew from 13 July 2005 that the Commissioner agreed with Alphapharm (because on that date the Commissioners notified Lundbeck A/S that it was proposed to amend the term of the extension to the 9 December 2012 date). Lundbeck A/S had not, in response, applied to extend time for an extension of term based on the registration of Cipramil or suggested that it may do so to any party in the Alphapharm proceedings. 264 I also accept that certain other matters would have been known to the parties to the Sandoz settlement agreement. They would have known that various generic entities, including Alphapharm, Arrow and Sandoz, were keen to launch their own escitalopram products and thus had an interest in seeing the 144 patent revoked (as sought by Alphapharm, Arrow and Sandoz in the Alphapharm proceedings until Sandoz discontinued its proceeding as provided for in the Sandoz settlement agreement). 265 It would have been known also that Lindgren J, on 21 June 2006, had ordered the various proceedings involving Alphapharm, Arrow and Sandoz, and Lundbeck A/S and Lundbeck AU, to be heard together in April to May 2007. While the parties could not know when Lindgren J would decide the case, or when any appeal might be heard and decided, it is likely that they would have anticipated knowing the outcome (including any appeal) before 13 June 2009. Considered objectively as at February 2007, the chance that they might not know the final outcome of the Alphapharm proceedings must have seemed remote. Considered objectively, the parties must also have recognised that litigation might be settled at any time, so there might never be a judicial determination in the Alphapharm proceedings. 266 Sandoz submitted that: Amongst other things, the Court was to determine the correct date of expiry of the Patent, with Lundbeck contending that it was 13 June 2014 based on the Lexapro Extension rather than 9 December 2012 as the Commissioner had found based on the inclusion of Cipramil on the ARTG, while Sandoz, Alphapharm and Arrow contended for 13 June 2009 based on the 20 year term of the Patent. 267 It may not matter, but Lindgren J was deciding whether and if so, how the Register should be rectified which would depend on the validity of the Commissioner's decision to extend the term of the 144 patent and then amend that extension (see, for example, [14]-[33] of Lindgren J judgment). The Register, as I understand it, reflected the initial extension date of 13 June 2014. Lindgren J had three competing contentions before him about the end of the term of the 144 patent, 13 June 2014, 9 December 2012, and 13 June 2009. Lundbeck contended that the correct date was 13 June 2014. The Commissioner was a party and contended that the correct date was 9 December 2012, as the Commissioner had decided on 19 May 2006: see, in particular, Lindgren J judgment at [32]. Arrow contended that the correct date was either 13 June 2009 or 9 December 2012: see Lindgren J judgment at [41]. Alphapharm contended that the correct date was 13 June 2009. This all forms part of the surrounding circumstances of the Sandoz settlement agreement. 268 Sandoz noted that Mr Sharkey decided that: …a good commercial outcome for Sandoz was to negotiate an arrangement with Lundbeck DN that would permit Sandoz to exit the litigation, while also providing it with commercial advantages in the form of a licence or distribution agreement. Sandoz would benefit regardless of which party won the litigation. 269 For present purposes, it is enough to say that given that three generic entities were trying to revoke the 144 patent, and one (Alphapharm) was also the subject of infringement proceedings, it would have been apparent to the parties to the Sandoz settlement agreement that Sandoz had little to gain from continuing in the proceedings and little to lose by trying to do a deal to exit the proceedings. The parties also would have appreciated that, for generic entities, every day counted for entry into a market to maximise the opportunity to take market share and so any early entry Sandoz could negotiate in exchange for discontinuing its proceedings would have value to Sandoz. And for Lundbeck A/S, one less party to deal with was not without value, but they still had to contend with Alphapharm (alleged to be infringing the 144 patent) and Arrow who were both seeking revocation of the relevant claims of the 144 patent. 270 Otherwise, I consider that Sandoz's principal construction involves an impermissible reliance on the subjective states of mind of the drafter of the Sandoz settlement agreement, Mr Sharkey, and the person with whom he was dealing at Lundbeck AU, its Managing Director Stefanus Maritz. For example, Sandoz submitted that: Mr Sharkey's evidence as to the surrounding circumstances and commercial purpose of the Settlement Agreement and the text of clause 3 of the Settlement Agreement was not challenged in cross-examination. In particular, there was no challenge to Mr Sharkey's evidence that the finding of the expiry date of the Patent in the Alphapharm Proceedings would determine Sandoz's effective start date of Sandoz's licences pursuant to the Settlement Agreement, or to his evidence as to the purpose of clause 3(1)(d). 271 In support of its principal construction, Sandoz stressed Mr Maritz's email of 25 January 2007 which said: As discussed I have now received feedback from my legal department and they have agreed to the following… … …How about offering Sandoz a 'free exit from the case' (as in the initial proposal) and then a 2 weeks early entry in either 2012 or 2015 [sic] depending on the outcome of the case. 272 It is common ground that "2015" is likely to be a typographical error and should be read as "2014". 273 Sandoz noted that this email indicates that "Lundbeck's internal lawyers had framed the agreement to be referrable to the outcomes of the Alphapharm proceeding". This is inaccurate. There was no agreement (yet). And it was not Lundbeck which drafted the agreement which came into existence, it was Mr Sharkey. Apart from this it may be accepted that this email discloses Mr Maritz's subjective understanding (and perhaps that of the lawyers with whom he conferred) that the Alphapharm proceedings would determine when the 144 patent expired. Mr Sharkey said that he understood that Lundbeck was proposing to allow Sandoz into the market before two of the possible dates that "Justice Lindgren might determine in the [Alphapharm] proceedings to be the expiry date of the" 144 patent. I do not consider this evidence of Mr Sharkey's (or Mr Maritz's) state of mind can be part of the relevant surrounding circumstances. 274 Mr Maritz's email referred to the "outcome of the case" whereas Mr Sharkey said he believed that the judgment of Lindgren J would determine the commencement of the licence. I do not accept that the prospect of an appeal from Lindgren J's decision was outside of the common contemplation of the parties at the time. Despite Mr Sharkey's evidence referring to Lindgren J's decision, Sandoz's principal construction assumed that the determination of the Alphapharm proceedings includes any determination in an appeal. Sandoz has not attempted to reconcile this inconsistency (perhaps because they are irreconcilable). On Mr Sharkey's evidence, if Lundbeck had failed before Lindgren J but succeeded in an appeal so that the extension of the term of the 144 patent until 13 June 2014 was valid, Sandoz's licence would have commenced on 31 May 2009 because Lindgren J had so determined (albeit wrongly on this hypothesis). I am not persuaded that any relevant objective circumstance can arise from Mr Sharkey's state of mind or, in any event, that Sandoz's principal construction accurately reflects Mr Sharkey's evidence.

9.5 The meaning of cl 3(1) 275 Sandoz's principal construction, as noted, is that the contingency in cl 3(1) ("if the Patent expires on") means, in effect, "if [it is determined in the Alphapharm proceedings, including any appeal, that] the 144 patent expires on" one or other of the nominated dates. 276 As noted, I accept that the Alphapharm proceedings are a critical part of the context within which the agreement was made. It could not be otherwise given that the agreement was settling Sandoz's part in the Alphapharm proceedings. So much is clear from the recitals, the definitions, and cll 1(a), (d), (e) and (f), 2(b), (c) and (d), 4 and 5. But this does not require Sandoz's principal construction to be accepted. 277 First, cl 3(1) makes no reference to the Alphapharm proceedings. It does not say that the licence is effective from the date of expiry as determined in the Alphapharm proceedings. Given the reference to Proceedings elsewhere in the Sandoz settlement agreement, it would have been easy to frame cl 3(1) in that way, but that is not how the clause is framed. 278 Second, and as also noted, I find it difficult to reconcile Sandoz's principal construction and its case that the licences commenced on 31 May 2009 (cl 3(1)) and 1 May 2009 (cl 3(2)). The Full Court did not publish reasons requiring Lundbeck's appeal to be dismissed until 11 June 2009. If, as Sandoz would have it, cl 3(1) means the determination of the expiry of the 144 patent in the Alphapharm proceedings including any appeal, then Sandoz could not have known on 1 or 31 May 2009 (when it says it acted upon the licence) that, in fact, it had a licence. While what occurred in 2009 is subsequent conduct and thus irrelevant to the task of construction, the relevant point is that Mr Sharkey's evidence about his state of mind, if relevant (which I think it is not), does not support Sandoz's principal construction. Sandoz cannot have it both ways. 279 Third, and again as noted, the parties must be taken to have contemplated that the Alphapharm proceedings might settle so there would be no decision in the Alphapharm proceedings. As a result, the contingency in cl 3(1), construed in accordance with Sandoz's principal construction, might never be satisfied. This is an objectively unlikely intention to attribute the parties. In contrast, I note, it was certain that the 144 patent would expire when it came to the end of its term, be that its initial or its extended term. In this context, making the licence depend on an unavoidable fact, that the 144 patent would expire at some time, appears an objectively likely intention of the parties compared to the intention which underlies Sandoz's principal construction. 280 Ultimately, however, it is the language of cl 3(1) which excludes Sandoz's principal construction. As I have said, the licence is a "licence to the Patent" (which is defined to mean the 144 patent) and is not a licence to exploit the invention the subject of the patent. The clause pre-supposes that the licence is to a patent which is in force and thus has not expired. The clause provides for the licence to be, in each case depending on the date of expiry of patent, for a period of two weeks before expiry. This is the context which, in my view, controls the meaning of the clause. 281 The words "if the Patent expires" describe a state or condition under the Patent Act. Within the context of that Act, a patent expires at the end of its term or, if its term is extended before the end of its term, at the end of its extended term. In the present case, the 144 patent had been extended before the end of its term, so that its term did not end until 13 June 2014 (subsequently amended to 9 December 2012). An issue in the Alphapharm proceedings was whether those decisions were lawful or not. This is where Lundbeck's construction and Sandoz's alternative construction come into play. Lundbeck would have it that "if the Patent expires on" in cl 3(1) means the ultimate expiry of the 144 patent pursuant to any extension of term, whether the extension is granted before or after the standard term of the 144 patent. This last qualification is critical because if an extension was granted only after 13 June 2009, the 144 patent would expire on 13 June 2009 irrespective of the subsequent extension of term. Accordingly, Lundbeck's case is that cl 3(1) should not be read as applying to the expiry of the 144 patent on 13 June 2009 because, on 25 June 2014, the term of the patent was extended to 9 December 2012. Sandoz's alternative construction (which Apotex and Aspen support) would have it that if the 144 patent reached the end of its term, as it did on 13 June 2009, then it expired and thus the licence under cl 3(1) commenced on 31 May 2009. 282 In support of its position Lundbeck submitted that any contrary construction would involve the conclusion that Lundbeck had contractually deprived itself of the possibility of obtaining an extension of term by a further application to the Patents Office. Lundbeck stressed that nothing in cl 3(1) or elsewhere in the agreement purported to prevent Lundbeck from making a further attempt to obtain an extension of term by application to the Patents Office. By contrast, Sandoz was restrained from commencing proceedings to revoke the 144 patent in cl 1(b). Further, what the parties bargained for was that Sandoz would get a licence for two weeks before the end of the term of the 144 patent. The 144 patent, Lundbeck said, was always going to reach the end of its term on 13 June 2009. The very subject of cl 3(1) is the extension of the term of the 144 patent as that alone explains the different dates. If the 144 patent reaching the end of its initial term on 13 June 2009 was intended by the parties to be the expiry of the patent as provided for in cl 3(1) then Sandoz would have a licence for two weeks and no more. 283 In response to Lundbeck's construction and supporting submissions, Sandoz submitted that: (1) The agreement "does not accommodate any future application for an extension of time and a second application for extension of term of the kind subsequently prosecuted by Lundbeck". (2) "To accommodate the grant of the Cipramil Extension Application after the expiry of the Patent, some form of words must be read into Clause 3(1) so as to provide that Clause 3(1)(a) would not operate, or to provide that the effective start date of the licence granted pursuant to Clause 3(1) would retroactively change, if Lundbeck DN were granted an extension of term after the Patent 'expires'". (3) "Lundbeck DN and Lundbeck AU were sophisticated parties with legal representation, and if they had wished to address the circumstances of that kind, could have done so expressly. Indeed, Lundbeck DN's lawyers made amendments to Clause 3(1) and chose not to make any such qualifications". (4) "…it is apparent from the surrounding circumstances of the Settlement Agreement that the parties were not contemplating those circumstances. To the extent the parties were contemplating an extension of term, it was the extension of term as determined in the Alphapharm Proceedings". 284 It was also noted for Sandoz that it is not the case that the 144 patent would necessarily reach the end of its term and expire on 13 June 2009. If a valid extension of term was granted before 13 June 2009 then the term of the 144 patent would be extended and the patent would not expire until the end of the extended term. Further, a licence may continue after a patent has expired, as s 145 of the Patents Act recognises. 285 A number of considerations weigh in favour of giving the words "if the Patent expires" the meaning for which the generic parties contend. 286 First, the natural and ordinary meaning conveyed by the words "if the Patent expires" is if the term of the 144 patent ends so the patent has expired. The term of the 144 patent ended on 13 June 2009 and the patent then expired. That the 144 patent so expired is a necessary pre-condition to the operation of s 79 of the Patents Act on which Lundbeck relies to bring these proceedings. For Lundbeck's construction to be correct it is necessary to read the words "if the Patent expires" as meaning if the patent reaches the end of its term and expires except if, at a time after the patent has reached the end of its term and expired, the term of the patent is extended. However, even if an extension of term was granted after 13 June 2009, no provision of the Patents Act provides that the 144 patent is taken not to have reached the end of its term and expired on 13 June 2009. To the contrary, s 79 depends on the term of the patent having expired and provides only for the vesting of rights during the period between expiry and the grant of an extension. While it is true that on the grant of the extension on 25 June 2014 the term of the 144 patent was extended until 9 December 2012, s 79 does not deem the patent not to have expired on 13 June 2009. The reality is that, as between 13 June 2009 and 25 June 2014, the 144 patent had expired. The only effect of the grant of the extension of term on 25 June 2014 was that s 79 was engaged to permit infringement proceedings to be commenced for acts between the expiry and the grant of the extension. 287 Second, it may be accepted that nothing in cl 3(1) or any other provision of the Sandoz settlement agreement excludes the possibility that Lundbeck A/S might apply to the Patents Office for another extension of term. However, I do not consider that this possibility drives the construction of cl 3(1). The clause is to be construed by reference to what it says, in context, and not by reference to what it does not say. 288 Third, an extension of term might be granted at any time after the end of the term of a patent. As a result, it might not be known for years if the patent had expired or not within the meaning of cl 3(1). It is objectively unlikely that the parties intended that to be the position when the trigger for the licence was to be an event (expiry of the 144 patent) known to be inevitable, at the apparent latest, by 13 June 2014. 289 Fourth, and contrary to Lundbeck's submission, it is not the case that the 144 patent would necessarily come to the end of its term and thus expire on 13 June 2009. As at the date of the Sandoz settlement agreement there were at least four circumstances in which the 144 patent would not expire on 13 June 2009, being: (1) In the Alphapharm proceedings Lundbeck succeeded in its challenge to the Commissioner's amendment to the extension of term of the 144 patent to 9 December 2012, and succeeded in defending the validity of the original extension of term to 13 June 2014. (2) Lundbeck failed in its case in the Alphapharm proceedings, but the Commissioner (or Arrow on its alternative case) succeeded in defending the validity of the Commissioner's decision to amend the extension of term of the 144 patent from 13 June 2014 to 9 December 2012. (3) The Court in the Alphapharm proceedings (either Lindgren J or on appeal) decided that the correct extension of term date was some date after 13 June 2009, but not 13 June 2014 or 9 December 2012. (4) Lundbeck A/S applied for an extension of time to apply for an extension of term to 9 December 2012 (or some other date) based on the Cipramil registration and, before 13 June 2009, the Commissioner validly granted that extension. 290 In contrast, as at February 2007, there was only one way in which the 144 patent might expire on 13 June 2009 and yet Lundbeck would obtain the right to sue for infringement for acts after that date - the possibility Lundbeck A/S might apply for an extension of time to apply for an extension of term of the 144 patent before 13 June 2009 based on the registration of Cipramil and be granted such an extension of term after 13 June 2009. 291 Lundbeck's construction treats this possibility, objectively remote as at February 2007, as if it controls the meaning of cl 3(1). For present purposes, hindsight is a disadvantage. We now know that Lundbeck A/S applied for an extension of time to apply for an extension of term based on the Cipramil registration on 12 June 2009, obtained an extension of time on 1 June 2011, and obtained an extension of term on 25 June 2014. But the Sandoz settlement agreement was executed in February 2007. At that time the primary focus of the parties, objectively ascertained, must have been the extension of term already granted and the amendment of that extension which were the subject of the Alphapharm proceedings. Moreover, to the extent that the possibility of it applying for an extension of time to apply for an extension of term was objectively known (which I have accepted), as noted, nothing in the Sandoz settlement agreement prevented Lundbeck from availing itself of that possibility and obtaining an extension of term based on the Cipramil registration before 13 June 2009. Had Lundbeck done so, the 144 patent would not have reached the end of its term on 13 June 2009 and thus would not have expired on that date within the meaning of cl 3(1). 292 It is necessary always to keep in mind that the Sandoz settlement agreement was executed in February 2007 and the earliest possible expiry date of the 144 patent was two and a half years away. I accept Lundbeck A/S was free to apply for an extension of time to apply for an extension of term of the 144 patent based on the registration of Cipramil if it chose to do so. Had it obtained such an extension before 13 June 2009 then, irrespective of the Alphapharm proceedings, I consider that the contingency in cl 3(1) would not have been satisfied. But what I do not accept is that as at February 2007 the parties should be taken to have meant by cl 3(1) that the 144 patent would not "expire" on 13 June 2009 despite it in fact doing so because Lundbeck A/S might obtain an extension of term after 13 June 2009. The effect of Lundbeck's case is that cl 3(1) should not be given its natural and ordinary meaning because of possibilities that, as at February 2007 when the Sandoz agreement was executed, must be objectively characterised as remote. 293 I do not consider it legitimate to test the operation of cl 3(1) against what actually occurred in the present case. As at February 2007 the parties could not have predicted what was going to occur between 2009 and 2014. 294 A difficulty with the clause is that the contingency "if the Patent expires" must always be a future contingency at the date of commencement of each licence as proposed because the licence was to commence two weeks before the contingency. Given the terms of cll 3(1) and 3(2) the parties must be taken to have understood that they would know by 1 May 2009 when the 144 patent would expire. This was a reasonable assumption but it left open the risk that they might not know if the contingency had been satisfied at the proposed licence commencement date(s). Both parties must be taken to have accepted this risk which, in February 2007, would have appeared insignificant. 295 For these reasons I consider that cl 3(1), in referring to "if the Patent expires on" means if the 144 patent reaches the end of its term and expires on one or other of the nominated dates. It does not mean if the 144 patent reaches the end of its term and expires on one or other of the nominated dates but excluding any case where, after the patent has reached the end of its term and expired, the term of the patent is extended. In the event, the 144 patent reached the end of its term and expired on 13 June 2009. The contingency in cl 3(1)(a) was satisfied and thus Sandoz had an irrevocable licence under cl 3(2) from 1 May 2009 and a licence under cl 3(1) from 31 May 2009. 296 However, this is not the end of the matter. What is the "Patent" to which Sandoz had a licence? In my view, it is the 144 patent before its expiry on 13 June 2009. As I have said, one thing is clear from cl 3(1). It is that the parties agreed that the period of the licence was two weeks before its expiry. As a result, there cannot be attributed to the parties any common intention that "the Patent" to which Sandoz was granted a licence is the patent the subject of the extension of term. To so conclude would be to re-write the deal the parties did. It is not to the point that cl 3(1) does not say that the licence expires on expiry of the patent. Nor does it matter that under s 145 of the Patents Act that a licence may be terminated on notice at any time "after the patent, or all the patents, by which the invention was protected at the time the contract was made, have ceased to be in force". What the parties bargained for was for Sandoz to receive a licence "to the Patent" for the period of two weeks before the patent expired. If, as I consider, the patent expired on 13 June 2009, the licence also must have ended on 13 June 2009. As I have said, Sandoz was not granted a licence to exploit the invention the subject of the patent. No doubt the licence "to the Patent" authorised such exploitation while the patent existed. The nature of the licence indicates, however, that it is tied to the existence of the patent before expiry. 297 A number of other considerations support this conclusion. 298 First, because the licence is expressed as a "licence to the Patent", the licence must be construed consistently with the balance of the clause. I am aware that "Patent" is a defined term in the settlement agreement, meaning "Australian Letters Patent Number 623144". I am aware also that it was this patent which was the subject of the grant of the extension of term. But it is apparent from the text and structure of cl 3(1) that the common intention of the parties was that Sandoz would have a licence for the period of two weeks before expiry of the patent. The clause pre-supposes that the "patent" is the 144 patent before it has expired. 299 Second, the word "Patent" can bear this consistent meaning throughout the settlement agreement without doing violence to its terms or objectively apparent intent. Indeed, it is only by giving this meaning to "Patent" that violence to the manifest common intention of the parties for Sandoz to have a royalty free licence for the period of two weeks before expiry of the patent can be avoided. 300 Third, as noted, it will be apparent that the licence does not involve any royalty. This makes sense in the context of the deal for Sandoz to have a licence for two weeks before expiry of the patent but makes no sense if Sandoz also thereby obtained a royalty free licence to exploit the invention if the term of the patent was extended after its expiry. 301 Fourth, if (as is the fact) the 144 patent expired on 13 June 2009, Sandoz's case that it had a licence to exploit the invention both before and after expiry of the patent and thus during the extended term of the patent once the extension had been granted does violence to the deal the parties did. It has the effect of transforming Sandoz's two week early entry licence under cl 3(1) into a royalty free licence for three years for the extended term of the 144 patent, which is inconsistent with the deal the parties bargained. It also gives "Patent" a meaning it cannot bear in the context of the settlement agreement, such as "the patent before it expires on 13 June 2009 and, if Lundbeck happens to obtain an extension of term after that date, the patent as subject to that extension of term". 302 Sandoz sought to avoid this result by emphasising that the licence granted by cl 3(1) is "an irrevocable non-exclusive licence to the Patent effective from" one of the nominated dates. This is immaterial. Clause 3(1), properly construed, controls both the commencement and the term of the irrevocable licence. The term of the licence is the period of two weeks before expiry of the patent. Within that two week period, the licence was irrevocable. The licence simply reached the end of its term on 13 June 2009. Accordingly, the status of the licence as irrevocable is beside the point. 303 For these reasons, I consider that Sandoz had a licence under the Sandoz settlement agreement. By cl 3(2) Sandoz had a licence to manufacture, import, market and offer to sell its escitalopram products from 1 May 2009. By cl 3(1)(a) it had a licence to sell its escitalopram products from 31 May 2009. Its licence reached the end of its term and ceased to operate on 13 June 2009 when the patent expired.

9.6 Was the licence varied on 28 May 2009? 304 Sandoz contended that, by a letter dated 28 May 2009, the parties agreed to vary the licence under the Sandoz settlement agreement. The background to the letter of 28 May 2009 is that on 24 April 2008 Lindgren J had rejected Lundbeck's arguments and found that the Commissioner wrongly granted the extension of term of the 144 patent until 13 June 2014 (as the time for applying for the extension of term had to be determined by reference to the registration of Cipramil on the ARTG). On 19 June 2008 Lindgren J ordered that the Register be rectified by removing the extension of term but stayed this order pending the appeal. The Full Court heard the appeal in February 2009 but had not yet delivered judgment. In the meantime, Sandoz had obtained approval for the listing of its escitalopram product, Esitalo, on the ARTG on 14 October 2008 and on the Pharmaceutical Benefits Scheme (PBS) on 1 May 2009. Unbeknownst to Lundbeck, Sandoz had also ordered and received from overseas supply of Esitalo on 27 May 2009. 305 On 26 May 2009 Lundbeck's lawyers wrote to Mr Sharkey of Sandoz noting the matters of which Lundbeck was aware and that the Sandoz settlement agreement "does not deal with the situation where the date on which the Patent will expire is not known with certainty". Lundbeck's solicitors proposed an exchange of undertakings. Mr Sharkey told Lundbeck's lawyers that Sandoz had already imported Esitalo. In the letter of 28 May 2009, undertakings were given. Sandoz undertook to Lundbeck A/S and Lundbeck AU that: 1. unless released from this undertaking by written notice from Lundbeck [Lundbeck A/S] or unless the finding of the Full Court of the Federal Court of Australia is to the effect that Australian Patent 623144 expires on 13 June 2009, it will not (without first providing Lundbeck at least 5 days written notice) promote, market, distribute, offer to sell, or sell any pharmaceutical product containing escitalopram. 306 Lundbeck A/S undertook to Sandoz that: 2. in the event that it is unsuccessful on the issue of the extension of term in Federal Court appeal (proceedings no.'s NSD 1048, 1052 and 1053 of 2008) it will pay damages (to be agreed or determined by order of a Court) to Sandoz for any disadvantage Sandoz may sustain by reason of its undertaking to our clients; and 3. once it has received notice of the date on which the Full Federal Court will deliver its judgment, Lundbeck will notify Sandoz as soon as possible of that date. 4. Lundbeck has also agreed that, pending the outcome of the Full Court appeal, Sandoz may import into Australia pharmaceutical products containing escitalopram on the condition that such importation is only into its own possession and not for further distribution. 307 It is not apparent to me why these cross-undertakings should be characterised as a variation to the licence under the Sandoz settlement agreement. On my construction of the Sandoz settlement agreement, and on the facts as found, Sandoz had a licence under cl 3(2) from 1 May 2009 (which excluded sale or supply) and a licence under cl 3(1) (which was unconfined) from 31 May 2009. Lundbeck was right in its letter of 26 May 2009 that the Sandoz settlement agreement "does not deal with the situation where the date on which the Patent will expire is not known with certainty". As such, Sandoz could not know if it did or did not have a licence until 13 June 2009 at the earliest. In the event, because the 144 patent reached the end of its term and expired on 13 June 2009, Sandoz did have a licence to do what it did in and leading up to 28 May 2009. 308 The undertaking did confine Sandoz's rights under cll 3(1) and 3(2) but the better characterisation is that the parties did what they said they were doing - they bound themselves by cross-undertakings in a manner which confined Sandoz's rights under the Sandoz settlement agreement. In so doing, they did not vary that agreement. They simply entered into another agreement, the effect of which was to confine Sandoz's rights to the extent specified in exchange for Lundbeck A/S's undertaking. 309 To the extent that Sandoz submitted that these events were relevant to the construction of the Sandoz settlement agreement, this cannot be so as they comprise conduct more than two years subsequently. Otherwise Sandoz submitted that the 28 May 2009 letter embodies a contract between the parties to the undertakings, Lundbeck A/S and Sandoz. I accept this submission. The cross-undertakings satisfy the legal requirement for a contract. The parties were bound by the terms they agreed in the cross-undertakings. 310 By 11 June 2009 the Full Court had found to the effect that the 144 patent expired on 13 June 2009 so Sandoz was no longer bound by its undertaking. 311 Sandoz contended that between 28 May and 11 June 2009 it acted in accordance with its undertaking and thus Lundbeck A/S's cross-undertaking in para 2 to pay damages is engaged. Lundbeck A/S undertook to pay damages to Sandoz "for any disadvantage Sandoz may sustain by reason of its [Sandoz's undertaking] to our clients [Lundbeck A/S and Lundbeck AU]". The fact that Sandoz did not promote, market, distribute, offer to sell, or sell any pharmaceutical product containing escitalopram between 28 May and 11 June 2009 does not prove that Sandoz sustained any disadvantage by reason of its undertaking. This is particularly the case given that Sandoz's undertaking was that it would not do those things without first providing at least five days' written notice. Sandoz did not suggest it gave such a notice between 28 May and 11 June 2009. 312 As a result, it was not sufficient for Sandoz to submit, as it did that: Pursuant to its obligations set out in the 28 May 2009 letter, Sandoz did not promote, market, distribute, offer for sell or sell Sandoz's escitalopram products between 28 May 2009 and the Full Court handing its judgment in the Alphapharm Proceedings on 11 June 2009. In addition, as set out above, Sandoz wrote to the Department of Health, notifying the Minister that it would be unable to guarantee supply of its Esitalo Products from 1 June 2009 until the Full Court in the Alphapharm Proceedings handed down its decision sometime between 1 and 13 June 2009. 313 Proof of these facts does not mean Lundbeck A/S could have had any liability to Sandoz under its undertaking. It is not apparent from these matters that Sandoz sustained any disadvantage by reason of Sandoz's undertaking. First, Sandoz did not know that the 144 patent would expire on 13 June 2009 until 11 June 2009 when, in any event, its undertaking was no longer in force. Second, and as is apparent from its subsequent conduct discussed below in the context of the claim for aggravated damages, Sandoz must be inferred to have had real doubt about the operation of the licence under cl 3(1) of the Sandoz settlement agreement because, even after 11 June 2009, it considered its launch of Esitalo to be at risk of infringing the 144 patent. Third, Sandoz was entitled to act, if it saw fit to do so, on five days' written notice but gave no such notice. From these matters I would infer that, irrespective of its undertaking, Sandoz would not have launched Esitalo before the Full Court's decision on 11 June 2009. As a result, it suffered no disadvantage by reason of its undertaking. 314 Otherwise, Sandoz contended that by reason of the cross-undertakings given on 28 May 2009, it had a licence to promote, market, distribute, offer to sell and sell products containing escitalopram from 13 June 2009 even if the 144 patent did not expire until 9 December 2012. I disagree. This argument seems to involve turning the negative stipulation in para 1 of the undertaking (Sandoz will not do things unless) into a positive grant of rights if the contingency ("unless the finding of the Full Court…") is satisfied. The reasoning which underpins this transformation is unsustainable. Lundbeck was not giving Sandoz a licence of any kind in para 1 of the undertaking. That paragraph merely confines Sandoz in the way specified and in exchange for the cross-undertakings that Lundbeck A/S gave in paras 2 to 4. In effect, the parties agreed that Sandoz would not exercise any rights under its possible two week licence "to the Patent" commencing on 31 May 2009 under cl 3(1) of the Sandoz settlement agreement in exchange for Lundbeck's undertaking to pay damages to Sandoz "for any disadvantage Sandoz may sustain by reason of it" having done so. I say "possible two week licence" because the parties did not know at the time whether any such licence existed or not (which may well explain why Sandoz was a willing participant to the exchange of undertakings by which it gave up an uncertain right for a certain right to damages provided, however, it could prove that it had sustained any disadvantage from giving its undertaking). 315 Paragraphs 2 to 4 also do not grant any form of licence to Sandoz. This is obvious for paras 2 and 3, but less obvious for para 4. The relevant facts are these. We know that Mr Sharkey told Lundbeck's lawyers that Sandoz had already imported it products into Australia. Mr Sharkey (but not necessarily Sandoz) believed the licence under cl 3(2) entitled this importation. Lundbeck's lawyers apparently disagreed. Paragraph 4 is nothing more than a recognition of a fact that existed. Paragraph 4 does not grant Sandoz any rights at all after the outcome of the Full Court appeal. Again, Sandoz's case appears to transform one thing (an agreement pending the outcome of the Full Court appeal) into another altogether (a licence to import after the outcome of the Full Court appeal). 316 Accordingly, I do not accept that the cross-undertakings themselves gave Sandoz a licence of any kind. They involved Sandoz agreeing not to exercise any rights which it might have had under cl 3(1) of the Sandoz settlement agreement, depending on the proper construction of that agreement.

9.7 The meaning of the releases in cll 2(b) and (c) 317 The terms of the Sandoz settlement agreement, including the releases in cll 1 and 2, are set out above. Sandoz contended that Lundbeck breached the releases in cll 2(b), (c) and (d) of the Sandoz settlement agreement by bringing these proceedings and denying the licences at various times. The extent to which Sandoz pressed these arguments given the abandonment of its cross-claim is unclear but I will deal with the issues as briefly as possible. 318 For the same reasons as set out above, it cannot have been a breach of any term of the Sandoz settlement agreement for Lundbeck to assert, as it did at various times, its construction of cl 3(1). 319 Otherwise, the key facts are that the "Proceeding" in cll 2(b), (c) and (d) is NSD 714 of 2006 commenced by Sandoz seeking revocation of the 144 patent. The "Proceeding" is not the proceedings as referred to in cl 1(a) of the Sandoz settlement agreement involving Alphapharm and Arrow. 320 It may be accepted, as Sandoz submitted, that a release can extend beyond that which is known to the parties at the time the release is granted: Doggett v Commonwealth Bank of Australia [2015] VSCA 351; (2015) 47 VR 302 at [63]. But the intention to do so and language used must be clear as, ordinarily, "general words in a release are limited to what was specifically in the contemplation of the parties at the time when the release was given": Karam v Australia & New Zealand Banking Group Ltd [2001] NSWSC 709 at [406]. 321 Sandoz submitted that: In the context of releases given in a patent dispute, the court's analysis of the commercial benefits of the settlement agreement may take into account the undoubted commercial benefits of the settlement agreement to the patentee of avoiding a challenge to the validity of the patent, and in so doing construing a release as extending to the subsequent extension term of the patent or a subsequent claim relating to the same patent. 322 So much may be accepted, but has nothing to do with the deal that the parties did in the Sandoz settlement agreement. R&J Lyons Family Settlement Pty Limited v 155 Macquarie Street Pty Limited [2006] NSWCA 177 does not assist Sandoz. In particular, Mason P's statement at [25], if anything, supports Lundbeck's construction of cll 2(b) to (d). Mason P said: The Settlement Deed was executed in the context of the imminent trial of the whole proceedings. It did not purport to lay to rest every conceivable difference between the parties, but its obvious intent was to end claims by the present claimants relating to the Easement proceedings they had launched. In my view "claims" in this context included claims, however framed, to rely upon the matters pleaded by the claimants in those proceedings as they stood at that time as the basis for some entitlement in their favour. The expression is not confined to existing claims or claims to a monetary remedy. 323 In circumstances where the releases are from claims, actions or causes of action "relating to the Proceeding", which was confined to Sandoz's claim that the 144 patent was invalid and should be revoked, it is impossible to construe the releases as preventing Lundbeck from asserting its construction of the Sandoz settlement agreement or seeking to enforce its rights as it believed them to exist given the terms of the Sandoz settlement agreement. Such matters have nothing to do with the "Proceeding" as defined in the Sandoz settlement agreement. 324 Sandoz's case also seems to have no regard to the exception in cl 2(d) which excludes from the release a claim, action or cause of action for breach or enforcement of the Sandoz settlement agreement. Consistent with business common sense, this demonstrates that the parties never intended that Lundbeck would not be able to assert and seek to vindicate its construction of cl 3(1). While Lundbeck's assertions and these proceedings are not for breach or to enforce the Sandoz settlement agreement they are founded on Lundbeck's construction of that agreement. On Lundbeck's primary construction of cl 3(1), Sandoz did not have a licence until December 2012 so that Sandoz's acts between 13 June 2009 and 9 December 2012 constituted an infringement of the 144 patent. Its alternative case is that if Sandoz had a licence from May 2009 because the 144 patent expired on 13 June 2009, then on the proper construction of cl 3(1) that licence expired on 13 June 2009. I have accepted Lundbeck's alternative case. Even if I had rejected all of Lundbeck's propositions, Sandoz's case involves concluding Lundbeck gave away its rights to assert its construction of cl 3(1), despite having expressly reserved to itself the right to claim for breach or enforcement of the agreement. In my view, this is untenable. 325 Sandoz's submissions do not confront the profound unlikelihood of it being the common intention of the parties that Lundbeck should not be able to assert and bring proceedings for infringement if, in Lundbeck's view, Sandoz acted as if it had a licence under the agreement when it did not. The fact that infringement of the 144 patent by Sandoz in the future would have been within the contemplation of the parties does not assist Sandoz, but Lundbeck. The purpose of the licence was so that Sandoz would not infringe the 144 patent if it exploited the invention in the two weeks before the patent expired. To attribute to the parties an intention to prevent Lundbeck from confining Sandoz to the licence Lundbeck thought it had granted Sandoz, in that context, is unsustainable. As Lundbeck submitted: It cannot seriously be suggested that these releases gave an open-ended release in respect of indeterminate future acts of infringement. This would lead to an absurd outcome in that Sandoz would have been free from the moment the Settlement Agreement was executed to start exploiting the Patent and would be wholly immune from allegations of infringement. To construe the releases in this way is a nonsense. 326 To the extent Sandoz relied on evidence about Mr Sharkey's state of mind, my observations above apply. Mr Sharkey's subjective intentions are not to be confused with the objective intention of the parties. 327 Sandoz also submitted that: Any construction of clause 2 of the Settlement Agreement that excludes proceeding NSD 647 of 2014 from its scope, begs the question what the release clause was otherwise releasing Sandoz from, and should be rejected. 328 The clauses are to be construed in the context as at February 2007. It must also be appreciated that Lundbeck's releases in cll 2(b), (c) and (d) reflect Sandoz's releases in cll 1(d), (e) and (f). It does not matter that the only possible claim relating to the Proceeding might have been a claim for costs with which the parties dealt expressly in each clause. If the clauses have that limited an operation, then so be it. It is no justification for attributing to the parties an intention which is irreconcilable with the fact that both parties reserved to themselves the right to enforce the agreement (cll 1(f) and 2(d)) so that it must have been their common intention that they would be able to assert and to litigate about their views of the licence granted. Lundbeck has done nothing more than this.

9.8 Sandoz - conclusions 329 For the reasons given above, Sandoz had a licence under the Sandoz settlement agreement as provided for in cl 3(1) (from 31 May 2009) and cl 3(2) (from 1 May 2009) until 13 June 2009. That licence provides no defence to Lundbeck's infringement case.

Further relevant facts 330 It is necessary to make some further findings of fact in order to resolve a number of other issues in dispute including Lundbeck's claim for additional damages under s 122 of the Patents Act, Sandoz's discretionary defence, and, to the extent it is pressed, its defence of innocent infringement under s 123.

10.1 Lundbeck 331 The following matters, which Apotex identified when it was still a respondent, must be accepted in respect of Lundbeck's conduct: (a) Lundbeck was on notice from 14 July 2005 of the Commissioner's view that CIPRAMIL contained (+)-citalopram for the purpose of s 70(3) and that accordingly the LEXAPRO extension had been wrongly granted. (b) Lundbeck was on notice from 14 July 2005, by reason of the letters sent by its Australian patent attorneys that it could, "in immediate defence of its position", apply for an extension of time in which to make an extension of term based on CIPRAMIL. (c) Lundbeck understood that if the Delegate decided to amend the Register pursuant to r 10.7(7) to record an expiry date of 9 December 2012 based on CIPRAMIL, then no application for an extension of time would be necessary. (d) Lundbeck challenged the validity of r 10.7(7) in Federal Court proceedings. (e) Justice Lindgren dismissed Lundbeck's challenge to the validity of r 10.7(7) on 1 March 2006: H Lundbeck A/S v Commissioner of Patents (2006) 150 FCR 269. (f) The Commissioner's Delegate, Dr Barker, determined on 19 May 2006 to amend the Register pursuant to r 10.7(7) to record an expiry date of 9 December 2012: Alphapharm Pty Ltd v H Lundbeck A/S (2006) 69 IPR 629. Lundbeck challenged that decision on appeal to the Federal Court, and did not seek in the alternative to preserve that outcome. (g) Justice Lindgren delivered judgment on Alphapharm's and Arrow's challenge to the extension of term on 24 April 2008, and ordered on that the extension of term ought to be removed from the Register: Alphapharm Pty Ltd v H Lundbeck A/S (2008) 76 IPR 618. On 19 June 2008 the Court so ordered. (h) Lundbeck took no step to notify Aspen of the possibility of an extension of time application until 2 February 2009 at the earliest. (i) In April 2009 (while the Full Court was still reserved), Corrs advised Lundbeck that it could, and should regardless of the outcome of the Full Court's decision, file an application for extension of time and term with the Patent Office: see GX TB, Ex 32, Tab 16. It suggested that this could occur in May 2009. (j) As at 26 May 2009, Lundbeck and its solicitors had the understanding, and expressly indicated when dealing with Sandoz, that the Full Court's decision would determine the "correct" term of the Patent. See the letter from Corrs to James Sharkey dated 26 May 2009 (CB Vol H4, Tab 269, p 4679): "Further, should the Full Court's decision be in Lundbeck's favour on the issue of extension of term, or should the orders disposing of the appeal be that the term be amended to record that the Patent will expire on 9 December 2012 (see below), it may be that Sandoz's conduct to date…will have been…an infringement of the Patent. […] We note that, once the Full Court's reasons are handed down, should Lundbeck be unsuccessful in its appeal in relation to the extension of term, it will seek an order that the extension of term not be removed but that the extended term of the Patent be amended to record that the Patent expire on 9 December 2012…We note that this was the result of the decision of the Delegate of the Commissioner of Patents which was the subject of appeal in Federal Court proceeding no. NSD 1078 or 2006. […] However, even if it was assumed that the Full Court's orders disposing of the appeal were that the extension of term be removed, with the result that generic pharmaceutical companies could enter the market from 14 June 2009…" (k) In a discussion between Lundbeck and Corrs on 26 May 2009, it was determined that Lundbeck should make the application to the Patent Office based on CIPRAMIL, and write to Apotex regarding the Patent Office extension of term if Lundbeck lost the extension of term in the Full Court: see GX TB, Ex 32, Tab 19. (l) As at 4 June 2009, Lundbeck and its solicitors had the understanding, and expressly indicated to Apotex, that an application would be made to the Full Court to record an extended term (to 9 December 2012) on the Register…. (m) In the event, Lundbeck took no step to notify Apotex of the possibility of an extension of time application until 12 June 2009, after the Full Court made its orders disposing of Lundbeck's appeal. (n) Although Lundbeck considered (on 11 June 2009) applying to the Full Court for an order amending the Register (exercising the powers of the Commissioner) to record an expiry date of 9 December 2012, it elected not to do so. (o) On 12 June 2009, Lundbeck undertook not to commence or threaten proceedings against any person for infringement of the Patent in respect of conduct engaged in after 13 June 2009. (p) Although Lundbeck had decided to make an application for an extension of time by the evening of 11 June 2009, it did not notify Aspen or Apotex of its intention to do so until after the Full Court made final orders on the appeal. (q) The extension of time application that Lundbeck made on 12 June 2009 was a "contingent" application in the sense that it was contingent on Lundbeck's application for special leave being unsuccessful. (r) Lundbeck's reason for filing the application on 12 June 2009 was that this was the last date upon which Lundbeck could make the application. (s) Lundbeck did not provide Apotex or Aspen with copies of the extension of time or term applications, or any supporting documents, or notify either of them of its undertaking not to sue or threaten to sue, on 12 June 2009. (t) In the course of prosecuting its extension of time application, Lundbeck gave assurances to the Commissioner, and undertakings to each of Aspen and Apotex that it would not seek injunctive relief against them. In its letter to the Commissioner confirming the formalised undertakings, Corrs acknowledged Lundbeck's representation during the hearing that it would not seek injunctive relief "so as to ensure that the [Products] remained available in the Australian marketplace" (emphasis added): see GX TB, Ex 32, Tab 36. (u) The High Court did not determine the question whether the Commissioner had power to extend time until 2014. In doing so, it divided 3:2. 332 These factual matters do not involve any impermissible impugning of facts found in Aspen Pharma Pty Ltd v Commissioner of Patents [2012] AATA 851; (2012) 132 ALD 648. I accept Lundbeck's argument that two facts as found were essential or indispensable to the outcome in that matter and that an issue estoppel thereby arises precluding the generic parties from putting in issue in this matter the following facts: 1. Lundbeck did not file an application for extension of term of the Patent based on the ARTG registration of Cipramil on 9 December 1997 prior to 12 June 2009 because of its erroneous view of the correct interpretation of the relevant provisions of the Patents Act 1990 (Cth) that any such extension could only be based on the ARTG registration of its product Lexapro which did not occur until 16 September 2003. 2. It was reasonable for Lundbeck to maintain its view as to the correct interpretation of the relevant provisions of the Patents Act 1990 (Cth) and to defer its application for an extension of term of the Patent based on Cipramil until at least the date of the decision of the Full Federal Court on 11 June 2009. 333 To the extent Sandoz submitted that no such estoppels exist, I disagree. It may be accepted that the Administrative Appeals Tribunal was dealing with an application for an extension of time. The power to extend time depended on Lundbeck having made an error which caused the delay. The Tribunal was satisfied that Lundbeck had made such an error. That finding, albeit in an administrative proceeding, gives rise to an issue estoppel binding the generic parties and Lundbeck who was joined as a party before the Tribunal. The Tribunal also had to decide what it considered to be the correct or preferable decision if the power to extend time existed: see [2012] AATA 851; (2012) 132 ALD 648 at [5]. In so doing the Tribunal had to decide whether to accept or reject the contention, recorded at [62], that "Lundbeck's conduct in the relevant period from 1999 was such as to disentitle it from being granted an extension". The Tribunal was satisfied that at all material times Lundbeck had acted reasonably. The findings to that effect were essential to the Tribunal concluding that there should be an exercise of discretion in Lundbeck's favour. Sandoz cannot now contend in these proceedings that it was not reasonable for Lundbeck to maintain its views about the correct interpretation of the Patents Act and to defer its application for an extension of term of the Patent based on Cipramil until at least the date of the decision of the Full Federal Court on 11 June 2009. 334 This said, I am also of the view that the facts asserted by Sandoz do not involve any assertion to the effect that Lundbeck did not hold an erroneous view about the relevant provisions of the Patents Act or that Lundbeck did not file its application based on Cipramil until 12 June 2009 because of any matter other than this erroneous view. They are a record of objective circumstances relevant to various issues which have to be resolved. They are relevant not because they impugn or tend to impugn the findings of error and reasonableness on which Lundbeck relies, but because they define some of the objective circumstances in which the infringements occurred. 335 In any event, it cannot be inferred from the evidence that Lundbeck deliberately delayed making an application based on the registration of Cipramil. It did not make such an application at an earlier time because it reasonably believed that would not accord with the statutory provisions. While it could have made an application at an earlier time based on a view of the provisions it believed to be wrong, it was not unreasonable for Lundbeck to act as it did in the circumstances. 336 Some further observations should be made which concern the objective circumstances. Lundbeck's views about the Lexapro registration being the relevant basis for an application for an extension of term of the 144 patent were not idiosyncratic. There is evidence that, in contrast to the course of events in Australia, extensions of term around the world were dealt with on the basis that the Lexapro registration was relevant, not the Cipramil registration. In these circumstances it was not unreasonable for Lundbeck to seek to sustain its view until the last day before expiry of the 144 patent. 337 The reasonableness of Lundbeck's conduct does not necessarily mean that Sandoz acted unreasonably. It too was confronted with circumstances which were extraordinary. While the capacity to apply for an extension of time to extend the term of the 144 patent existed for so long as the patent remained in force, the litigious efforts which had been the focus of so much energy before 12 June 2009 may well have led a reasonable person in the position of the generic parties to believe that the course of that litigation was likely to be determinative. Given that the extension of time required was in the order of 10 years it would not have been unreasonable for any person in the position of Sandoz to assume that, while Lundbeck could apply for such an extension, it may well not do so or that, once that application was made, it may well be unlikely to succeed. 338 This is not to say that, for these reasons or because Lundbeck had undertaken to the Commissioner and the Court not to sue or threaten to sue any person for infringement pending the determination of its special leave application (noting that the extension of term until 13 June 2014 remained on the Register pending the special leave application being determined), Sandoz is to be taken to have assumed anything other than that Lundbeck would be likely to do all in its power to maximise its rights, including to claim damages for infringement if it could possibly do so. As soon as the application for the extension of time was made on 12 June 2009, Lundbeck's sole purpose, as Sandoz must have known, was to enable Lundbeck to sue for damages for infringement for the period between 13 June 2009 and 9 December 2012. But as Sandoz also must have known, Lundbeck had to clear a number of significant hurdles before it could sue for damages for that period and that there was no guarantee Lundbeck would secure the rights to enable it to do so.

10.2 Sandoz 339 Insofar as Sandoz is concerned, the position can be summarised shortly. My conclusions are that while Mr Sharkey believed that Sandoz had the benefit of a licence, he was not the relevant decision-maker. The relevant decision-makers from the parent company Sandoz AG, who were based overseas, must be inferred not to have shared Mr Sharkey's view that there was a licence which authorised Sandoz's conduct. To the contrary, they considered the launch to be at risk of infringement. If FFC#1 had been decided against the generic companies, then Sandoz would have been forced to confront the issue of the efficacy of its licence at levels above Mr Sharkey, but this was not required because Lundbeck's appeal was dismissed. Accordingly, I do not accept Sandoz's submission that it acted on the basis of a belief that it had a licence which authorised its activities. 340 Further, to the extent steps had been taken in Australia to launch without the approval of these decision-makers it is apparent that, internally, this was considered to be in breach of Sandoz's protocols. Sandoz could have decided not to launch its products at any time. Ultimately, it decided to launch for commercial reasons having weighed the legal risks to it against the commercial opportunities. No doubt, the licence formed part of this assessment but it was plainly not considered a reliable defence by the relevant decision-makers. 341 Sandoz knew as early as 27 May 2009, by reason of its own legal advice, that Lundbeck could and may apply for an extension of time to apply for an extension of term based on Cipramil registration. It knew also that, if Lundbeck succeeded, then but for the potential operation of the licence, it would be at risk of a claim for damages. Sandoz acted because it assessed the commercial rewards to be worth the risk.

10.3 General matters 342 I do not accept that Sandoz acted unreasonably in opposing the applications for the extension of time and term or by denying infringement in the present case. Just as Lundbeck was entitled to seek to maximise its legal rights, the generic parties were entitled to protect their own positions. All entities involved appear to have taken every opportunity, at every stage, to protect their own positions. None is in a position to criticise another for doing so. There were real and reasonably available arguments for the generic parties to put at every stage of the process, including in this matter. They did not act unreasonably in doing nothing different from what Lundbeck has done at every stage.

Innocent infringement and discretionary defence 343 To the extent that Sandoz submitted that no award for damages should be made against it on the basis of s 123 of the Patents Act or in the exercise of discretion, I disagree. 344 Section 123(1) provides that: A court may refuse to award damages, or to make an order for an account of profits, in respect of an infringement of a patent if the defendant satisfies the court that, at the date of the infringement, the defendant was not aware, and had no reason to believe, that a patent for the invention existed. 345 It may be accepted that, because the 144 patent expired on 13 June 2009 and did not obtain the benefit of s 79 until 25 June 2014 it was a fact that no patent for the invention existed. In that sense, Sandoz could not have been aware and had no reason to believe that a patent for the invention existed. The discretion in s 123, accordingly, is available. 346 However, this is not an appropriate case for the exercise of the discretion. Ultimately, the fact that must be given determinative weight is that, no matter the confusion about the licence and the unusual circumstances, the evidence discloses that Sandoz made a calculated commercial decision to launch its products at risk. It knew that Lundbeck had applied for the extension of time and term. It knew that the only purpose of Lundbeck so doing was to be able to bring infringement proceedings against generic entrants to the market which Lundbeck alleged were infringing is patent, the purpose not being to claim damages and/or an account of profits. As Lundbeck set out it in its final submissions: …the short answer to this argument is that Sandoz was well aware of the existence of the patent and the possibility of an extension of time and term. Sandoz' launch was an "at risk" launch. Sandoz internally appreciated that the launch on 15 June 2009 was at risk even though Sandoz was aware of the terms of the Settlement Agreement. For example, Ex 17 demonstrates that Sandoz recognised that the start date of its licence was still uncertain because of the fact that Lundbeck might get an extension of time and then term. Further, Sandoz recognised that if that occurred, it would be liable for damages or an account of profits for patent infringement, and, in effect, nothing in the Settlement Agreement would avoid that outcome (see p3/8 penultimate paragraph, p 5/8 last paragraph). Ex 14 is to similar effect and is dated 17 June 2009, which is after the end of the original term of the patent (see p 0016_1, p 0016_5). Mr Turner, the witness that Sandoz put forward for the present hearing, confirmed in his recent cross-examination that the launch was at risk. 347 The same answer must be given to any argument that there should be an exercise of discretion against awarding damages. It may be accepted that s 122(1) of the Patents Act provides that: The relief which a court may grant for infringement of a patent includes an injunction (subject to such terms, if any, as the court thinks fit) and, at the option of the plaintiff, either damages or an account of profits. 348 While I prefer Lundbeck's argument that damages are not a discretionary remedy (in contrast to the remedies of the grant of an injunction and/or an account of profits), not much more need be said about this issue other than that, if there is any discretion, I would not exercise it in the circumstances of this case in favour of Sandoz. Sandoz made its own commercial decision in its own commercial interests. It took a calculated risk. It must be inferred that it received the financial benefit of having done so. The unusual circumstances of the case do not alter the fundamental fact that Sandoz knew that Lundbeck was seeking to ensure it could sue generic entrants to the market before 9 December 2012 for damages or an account of profits and took a risk that Lundbeck would not obtain the rights it needed to do so. Sandoz turned out to be wrong in the calculated risk it took. In the face of these fundamental facts, there is no basis to exercise any discretion not to award damages against Sandoz (assuming any such discretion exists). 349 As to the existence of the discretion, I note Lundbeck's submission that the proposition is implicitly inconsistent with the Full Court's judgment in Knott Investments Pty Ltd v Winnebago Industries, Inc (No 2) [2013] FCAFC 117; (2013) 305 ALR 387 in which the applicant was denied an account of profits due to unreasonable delay but the question of damages was remitted to the trial judge. On the remittal, in Winnebago Industries Inc v Knott Investments Pty Ltd (No 4) [2015] FCA 1327; (2015) 241 FCR 271 Yates J said: 65. First, the respondents focused on the restitutionary aspect of damages awarded under the user principle. Relatedly, they pointed to the Full Court's refusal to remit the applicant's claim for an account of profits. The respondents then argued that, by seeking damages based on the user principle, the applicant was seeking to "outflank" the Full Court's refusal to remit the claim for an account of profits by a "doctrinal sleight of hand" which involved the applicant seeking, in substance, nothing more than an account of profits calculated by an expense notionally saved by the respondents - a reasonable royalty. Thus, the respondents argued, the applicant was effectively seeking a remedy already denied to it by the Full Court. 66. I do not accept this submission. Although damages awarded under the user principle have an avowedly restitutionary aspect, it is a conceptual error to confuse damages in tort with the separate remedy in equity that compels a wrongdoer to give up profits which, having been derived by wrongdoing, should not be retained. Plainly, in the present case, the Full Court's refusal on discretionary grounds to remit the applicant's claim for an account of profits was informed by the equitable nature of that relief. But neither matter relied upon by the Full Court in refusing to remit the applicant's claim for an account of profits stands as a reason for refusing the applicant's claim for damages. Delay is no bar to a claim for damages, other than by preclusion and extinguishment of the underlying action by statute. And a defendant seeking to resist a claim for damages calculated under the user principle (here, by reference to a reasonable royalty or licence fee) is not confronted or vexed with the difficulties that might confront and vex a defendant seeking to resist an order to account for and give up profits wrongfully made. Indeed, had that not been so, the Full Court would not have remitted the applicant's claim for damages. It is inconceivable that, having refused to remit the applicant's claim for an account of profits for stated discretionary reasons, the Full Court would have immediately put out of mind those reasons when remitting the applicant's claim for damages. 350 Unilin Beeher BV v Huili Building Materials Pty Ltd (No 2) [2007] FCA 1615; (2007) 74 IPR 345, to which Sandoz referred, is immaterial. Allsop J (as he then was) was considering a claim for an account of profits which is a discretionary remedy, not a claim for damages.

Additional damages? 351 This all said, I also do not find persuasive Lundbeck's arguments that the circumstances of this case justify an award of additional damages having regard to s 122(1A) of the Patents Act. 352 Section 122(1A) provides that: (1A) A court may include an additional amount in an assessment of damages for an infringement of a patent, if the court considers it appropriate to do so having regard to: (a) the flagrancy of the infringement; and (b) the need to deter similar infringements of patents; and (c) the conduct of the party that infringed the patent that occurred: (i) after the act constituting the infringement; or (ii) after that party was informed that it had allegedly infringed the patent; and (d) any benefit shown to have accrued to that party because of the infringement; and (e) all other relevant matters. 353 Although s 122(1A) of the Patents Act has received relatively little judicial attention since it was introduced in 2007, authorities relating to s 115(4) of the Copyright Act 1968 (Cth), s 75(3) of the Designs Act 2003 (Cth) and s 126(2) of the Trade Marks Act 1995 (Cth) are relevant by analogy: Zetco Pty Ltd v Austworld Commodities (No 2) [2011] FCA 848 at [261]-[269] (Bennett J), applied in Industrial Galvanizers Corporation Pty Ltd v Safe Direction Pty Ltd [2018] FCA 1192 at [131]-[133] (Burley J), Pacific Enterprises (Aust) Pty Ltd v Bernen Pty Ltd [2014] FCA 1372; (2014) 321 ALR 715 at [13]-[14] (Pagone J) and Australian Mud Company Pty Ltd v Coretell Pty Ltd (No 4) [2015] FCA 1372 at [426]-[428] (McKerracher J); see also Truong Giang Corporation v Quach [2015] FCA 1097; (2015) 114 IPR 498 at [129]-[140], cited with approval in Geneva Laboratories Ltd v Prestige Premium Deals Pty Ltd (No 5) [2017] FCA 63; (2017) 122 IPR 279 at [78] and [80]. 354 In Truong Giang at [129]-[140] Wigney J described eight principles applicable to additional damages under intellectual property legislation, which Bromwich J summarised in Geneva Laboratories at [80] as follows: (1) there is no need for proportionality to compensatory damages (although as noted below there is nothing to indicate that that is a forbidden consideration or benchmark); (2) an award involves an element of penalty; (3) judicial disapproval has a part to play; (4) the terms of the statutory provision as to the non-exhaustive matters that can be taken into account in considering an award of additional damages do not constitute a precondition for them to be awarded; (5) flagrant conduct includes that which is deliberate and calculated in disregard of the injured party's rights, or a cynical pursuit of benefit; (6) post-infringement conduct such as in the conduct of the proceedings is relevant more to costs than to additional damages but not irrelevant; (7) additional damages encompass but are not the same as aggravated or exemplary damages at common law; and (8) specific deterrence has a part to play, including general deterrence. 355 At least one more principle should be added, the purpose of stripping the infringer of "all pecuniary benefits received from the infringement": Aristocrat Technologies Australia Pty Ltd v DAP Services (Kempsey) Pty Ltd [2007] FCAFC 40; (2007) 157 FCR 564 at [48]. 356 While I accept that Sandoz's infringement was calculated, it was not flagrant. In Futuretronics.com.au Pty Ltd v Graphix Labels Pty Ltd (No 2) [2008] FCA 746; (2008) 76 IPR 763 at [19] Besanko J said that: The meaning of "the flagrancy of the infringement" has been described in various ways. In Ravenscroft v Herbert and New English Library Limited [1980] RPC 193 Brightman J referred to it as "scandalous conduct, deceit and such like; it includes deliberate and calculated copyright infringement"; in Prior v Lansdowne Press Pty Ltd [1977] VR 65 at 70, Gowans J (at 70) referred to it as a "calculated disregard of the plaintiff's rights, or cynical pursuit of benefit"; and in Raben Footwear Pty Ltd v Polygram Records Inc (1997) 75 FCR 88 at 103 flagrancy was said to connote conduct which could probably be described as "glaring, notorious, scandalous" or "blatant". In Polygram [Pty Ltd v Golden Editions Pty Ltd (No) 2 (1997) 38 IPR 451] at 461-462 Lockhart J said that flagrancy did not include mere mistakes or carelessness and that if the infringer mistakenly believed that he or she owned the relevant copyright, or acted in the bona fide belief that no copyright subsisted in the plaintiff's work then the conduct was not flagrant. 357 Lundbeck did not have the 144 patent when Sandoz infringed. Lundbeck had contingent rights but the contingencies (the extensions of time and term) were uncertain. Lundbeck itself could not have known whether it would succeed or not in obtaining the extensions it required in order to be able to avail itself of the rights provided by s 79. 358 I do not consider that the "need to deter similar infringements of patents" arises on the facts of the present case. The circumstances in which Sandoz's infringements occurred are unusual, perhaps even unprecedented. I accept that a person in Sandoz's position in 2009 was not entitled to assume that Lundbeck would be unable to satisfy the contingencies to give it rights to enforce the 144 patent under s 79 of the Patents Act. This I why I have described Sandoz's conduct as involving a calculated risk. This said, however, the "need to deter similar infringements of patents" must fall within the lower end of the spectrum of need. 359 While Lundbeck always made plain to Sandoz that it would seek to secure its patent rights by applying for the extensions of time and term and would hold Sandoz liable for infringement, apart from taking the calculated commercial risk which it took, there was no particular conduct of Sandoz which would justify the award of additional damages. 360 Sandoz must have obtained the benefit of its infringements in the form of net profits on its sales of the infringing products. This benefit, however, can be adequately accounted for in the form of an award of damages without an award of additional damages. 361 Otherwise, while I have rejected Sandoz's contention that its conduct was carried out on the basis of a belief that it had the benefit of the licence from Lundbeck, I accept that it is relevant that the uncertainty about the status of the 144 patent between 13 June 2009 and 9 December 2010 was not the result of Sandoz's conduct. Sandoz was entitled to oppose Lundbeck's various attempts to enforce the patent and secure an extension of term of the patent, just as Lundbeck was entitled to take all steps it could to secure the extension. 362 I also reject Sandoz's submission as follows: The Court should not seek to deter other generic pharmaceutical companies from adopting the same conduct as Sandoz, in circumstances where an extension to a pharmaceutical patent is extended well after the expiry of the standard patent term. Rather, the Court should endorse, not condemn, Sandoz's conduct in preparing to launch, launching and continuing to sell and offering to sell pharmaceutical products in connection with judicial findings as to the expiry of the standard patent term in the Alphapharm Proceedings, and on its contractual licence agreements. 363 The problem with this submission is that it fails to recognise that the enforcement of Lundbeck's rights against Sandoz in the circumstances of this case is a result of the statutory scheme. That is, the statute permits an extension of time and term to be granted, and s 79 thereby to be engaged, after a patent has expired. As a result, for the Court to "endorse" Sandoz's conduct in taking a calculated commercial risk that Lundbeck would not be able to satisfy the relevant contingencies would be contrary to the statutory scheme. There can be no endorsement of such conduct. This does not negate the fact, however, that the circumstances of Sandoz's infringement were unusual (to say the least) and I see nothing which would attract the opprobrium of the Court such as to make an award of additional damages appropriate. 364 Lundbeck also said that "the manner in which Sandoz has elected to conduct these proceedings, including its pursuit of an untenable Cross-claim which it abandoned shortly after it opened that case" justified an award of additional damages. I do not accept this submission. It is true that Sandoz abandoned its cross claim at an early stage. But there was nothing in Sandoz's conduct that was "high-handed", "dishonest" or "recalcitrant" and it did not conduct itself so as to render compensatory damages inadequate. Such factors have been seen as potentially calling for an award of additional damages: see eg Truong Giang at [138] citing Facton Ltd v Rifai Fashions Pty Ltd [2012] FCAFC 9; (2012) 199 FCR 569 at [44] (Lander and Gordon JJ) and [69] (Gilmour J); Review Australia Pty Ltd v New Cover Group Pty Ltd [2008] FCA 1589; (2008) 79 IPR 236 at [53] (Kenny J). The short-lived cross claim was an ordinary incident of litigation. It can be dealt with by an order for costs in the ordinary course.

Infringement and liability for damages 365 For the reasons given above Sandoz infringed the 144 patent by its supply of its escitalopram products between 15 June 2009 and 9 December 2012, being the extended term of the 144 patent. None of Sandoz's defences to the claims against it for damages under s 122(1) has been established. This said, there are other issues concerning both the capacity of the various Lundbeck entities to make claims for damages against Sandoz and quantification of their losses which I now must resolve.

Damages for patent infringement - principles 366 The general principles applying to the assessment of damages for patent infringement were not in dispute. 367 In Generic Health Pty Ltd v Bayer Pharma Aktiengesellschaft [2018] FCAFC 183 the Full Court said at [147]: The purpose of an award of damages for patent infringement is compensatory, rather than punitive: Bailey v Namol Pty Ltd [1994] FCA 733; 53 FCR 102 at 110-111 and the other cases collected in Seafolly Pty Ltd v Fewstone Pty Ltd [2014] FCA 321; 313 ALR 41 at [506]. The appropriate quantum of damages is that which will restore the patentee to the position it would have been in had the infringement(s) not occurred: General Tire and Rubber Company v Firestone Tyre and Rubber Company Limited (1976) 93 RPC 197… and the other cases cited in Pacific Enterprises (Aust) Pty Ltd v Bernen Pty Ltd [2014] FCA 1372; 321 ALR 715 at [5]. 368 The party claiming damages bears the onus of proving loss: Aristocrat Technologies Australia Pty Limited v DAP Services (Kempsey) Pty Limited (in liquidation) [2007] FCAFC 40; (2007) 157 FCR 564. 369 As also explained by the Full Court in Generic Health at [181] and [182], the question of the sales that a party whose patent rights have been infringed would have made but for the infringements of the patent involves a past hypothetical event. In this sense, what the party whose rights have been infringed has lost is an opportunity to sell its products or as much of its products, at the price it would have sold them for, but for the infringing conduct. On that basis, the value of the lost opportunity is to be assessed "by reference to the degree of probabilities or possibilities" involved in the hypothetical counterfactual: Sellars v Adelaide Petroleum NL [1994] HCA 4; (1994) 179 CLR 332 at 355. 370 The Full Court in Generic Health continued as follows: 183 Another way to express the same point is as Brennan J described it in Sellars at 368: Although the issue of a loss caused by the defendant's conduct must be established on the balance of probabilities, hypotheses and possibilities the fulfilment of which cannot be proved must be evaluated to determine the amount or value of the loss suffered. Proof on the balance of probabilities has no part to play in the evaluation of such hypotheses or possibilities: evaluation is a matter of informed estimation. 184 This theme of informed estimation resonates with some observations made by Hayne J in Placer (Granny Smith) Pty Ltd v Thiess Contractors Pty Ltd [2003] HCA 10; 196 ALR 257 at [37] and [38]. Any estimation must be done "with as much precision as the subject matter reasonably [permits]". Mere difficulty in estimating damages does not relieve a court "from the responsibility of estimating them as best it can". 185 Interestingly, Staughton LJ in Gerber Garment Technology [1997] RPC 443 at 459 cited Sellars with approval and discussed the matter in terms of evaluating the loss of a chance. Gerber Garment Technology both at first instance ([1995] RPC 383 per Jacob J at 407 and 408) and on appeal ([1997] RPC 443 per Staughton LJ at 459 and 460) evaluated and applied a loss of a chance approach based on the possibilities and probabilities. Moreover, Jacob J at [1995] RPC 383, 395 also invoked the themes of Lord Diplock in Mallett v McMonagle [1970] AC 166; [1969] 2 WLR 767 at 176, which resonate with the themes discussed in Malec [Malec v J C Hutton Pty Limited [1990] HCA 20; (1990) 169 CLR 638] at 639 and 643 that we have referred to above. 186 Whichever way one expresses it, in assessing the possibilities or probabilities of a hypothetical counterfactual, one is engaged in the task of estimation, even if the estimation involves an assessment of the counterfactual as being close to a certainty. But being close to a certainty is not the same thing as a certainty. If one is estimating, one still needs to apply a discount, albeit a very modest one, to reflect the assessment that one is not at a certainty. If one is looking at the value of a lost opportunity which is not certain to occur, then the valuation must involve some discount, even if a very modest one. 187 Given that Generic Health is a wrongdoer, it may be accepted that damages should be liberally assessed, although not, of course, to punish it: General Tire (1976) at 212 per Lord Wilberforce [General Tire & Rubber Company v Firestone Tyre & Rubber Company Limited (1975) 92 RPC 203]. The object is to compensate Bayer. To say that damages should be liberally assessed in no way cuts across what we have just said: that, in estimating or valuing a lost opportunity or in assessing a hypothetical counterfactual for any scenario short of certainty, some discount must be made to reflect that less than certain position, even if the discount is very modest indeed. 188 Also, whilst the "five-stage approach" in Norm Engineering [Norm Engineering Pty Ltd v Digga Australia Pty Ltd [2007] FCA 761; (2007) 162 FCR 1] may be helpful in some circumstances to order analysis, it should not be seen as a definitive test for what is an evaluative assessment on all the available evidence. 371 In Malec v J C Hutton Pty Limited [1990] HCA 20; (1990) 169 CLR 638 at 643 Deane, Gaudron and McHugh JJ explained that: If the law is to take account of future or hypothetical events in assessing damages, it can only do so in terms of the degree of probability of those events occurring. The probability may be very high - 99.9 per cent - or very low - 0.1 per cent. But unless the chance is so low as to be regarded as speculative - say less than I per cent - or so high as to be practically certain - say over 99 per cent - the court will take that chance into account in assessing the damages. Where proof is necessarily unattainable, it would be unfair to treat as certain a prediction which has a 51 per cent probability of occurring, but to ignore altogether a prediction which has a 49 per cent probability of occurring. Thus, the court assesses the degree of probability that an event would have occurred, or might occur, and adjusts its award of damages to reflect the degree of probability. The adjustment may increase or decrease the amount of damages otherwise to be awarded…The approach is the same whether it is alleged that the event would have occurred before or might occur after the assessment of damages takes place. 372 In Norm Engineering Pty Ltd v Digga Australia Pty Ltd [2007] FCA 761; (2007) 162 FCR 1 Greenwood J proceeded on this basis: 266. In the circumstances, an approach to determining an assessment of reasonable compensation, might be to adopt a stepped method which provides a reasoned foundation for an assessment of compensatory loss. The steps might be these. 267. First, examine the number of sales made by the respondent of the relevant 4 in 1 buckets. 268. Secondly, assume that since Digga was trying to meet the market leader by introducing its 4 in 1 bucket, it was at least trying to capture sales from the supply side market leader (and no doubt from other suppliers). 269. Thirdly, assume that the number of sales actually made by the respondent represents sales, on balance, the applicant would have made, as a starting point. 270. Fourthly, in recognition that it does not follow that a sale by Digga would necessarily have been a sale by the applicant, a discount factor must be applied either to the number of sales or to the amount of compensation determined according to proper principles, so as to reflect a proportionate sense of the impact upon the applicant of the respondent's conduct of selling a competing supply side substitute incorporating the infringing component. 271. Fifthly, a further discount factor needs to be taken into account to reflect the particular circumstances of the conduct. 373 Otherwise, as Sandoz submitted: (1) Loss caused by lost sales is best represented by net lost profits, meaning "revenue less all costs including variable and indirect costs, but not including income tax": TS & B Retail Systems Pty Ltd v 3Fold Resources Pty Ltd (No 3) [2007] FCA 151; (2007) 158 FCR 444 at [207]. (2) "…the calculation of damages [is] not an exercise required to be undertaken with mathematical precision and that the Court must do its best on the evidence before it": Advanced Building Systems Pty Ltd v Ramset Fasteners (Aust) Pty Ltd [2001] FCA 1098; (2001) 52 IPR 305. (3) Damages must be proved with a degree of precision which reflects the proof that is reasonably available to the parties and, if the evidence called on behalf of a patentee does not provide a foundation for a proper estimate of damages, the Court should decline to make one: Placer (Granny Smith) Pty Ltd v Thiess Contractors Pty Ltd [2003] HCA 10; (2003) 196 ALR 257 at [38]. It is not enough merely to show "wrongful conduct": Aristocrat Technologies Australia Pty Limited v DAP Services (Kempsey) Pty Limited (in liquidation) [2007] FCAFC 40; (2007) 157 FCR 564 at [35] and [103].

The general approach to quantifying loss for patent infringement 374 For Lundbeck, Owain Stone, forensic accountant, calculated loss. For Sandoz, Tony Samuel, accountant and valuer, re-calculated loss using the same model (which was agreed) but used some different inputs. Mr Samuel also gave evidence contesting some of the claims as a matter of principle. 375 In its final submissions Lundbeck provided the following explanation of the general state of the accounting evidence which I adopt. 376 Lundbeck said: First, Mr Stone identifies the following forms of loss as the basis for his loss calculations: (a) Lundbeck AU: lower prices for LEXAPRO (but no change in the volume of sales of LEXAPRO); (b) CNS Pharma: lower sales volumes of ESIPRAM and lower prices for ESIPRAM; (c) Lundbeck DK: loss of gross margin on sales by it to CNS Pharma of ESIPRAM and a lower year-end transfer of funds from Lundbeck AU as a result of lower gross margins earned by Lundbeck AU and CNS Pharma. There does not appear to be a dispute about the identification of these potential forms of loss. It is also noteworthy that Lundbeck's approach is conservative. It assumes that all sales lost by it were sales of ESIPRAM, rather than a mixture of ESIPRAM and LEXAPRO. This leads to a lower loss calculation and simplifies the assessment of loss by avoiding the need to estimate the split between ESIPRAM and LEXAPRO. Secondly, Mr Stone summarises the methodology he has used to determine his loss calculations as follows: The estimation of loss and damage therefore requires the identification, and then comparison, of the financial outcomes associated with the following: (a) the position which the claimant is in as a result of the alleged infringement ('Actual Position') and the resulting cash flows ('Actual Cash Flows'); and (b) the position which the claimant would have been in had the alleged infringement not occurred ('Counterfactual Position') and the cash flows that would have resulted ('Counterfactual Cash Flows'). Estimation of loss and damage requires that estimates be made of the timing and amount of all relevant cash flows, including revenues, expenses, capital expenditure, movements in working capital, and tax. In my experience, these cash flows are generally assessed on a periodic basis, with monthly periods or annual periods commonly adopted. Leaving aside some matters of detail, there does not appear to be a dispute between the parties as to the appropriateness of this approach. Thirdly, Mr Stone's calculations are based on two periods: an initial period and a springboard period. (a) The initial period ran from 15 June 2009 (when the generic parties started to enter the market) to 31 December 2012. This period continued slightly beyond the extended term of the patent because 1 January 2013 was the first PBS listing date that would have been available to generic parties. The vast majority of the loss claimed by Lundbeck falls within the initial period. (b) The springboard period commenced on 1 January 2013. As at that date, as a result of its allegedly infringing conduct, Sandoz had an established position in the marketplace which it otherwise would not have had at that time. Further, the allegedly infringing conduct had already resulted in the price of both LEXAPRO and ESIPRAM being reduced. Accordingly, to take account of the springboard period, Mr Stone quantifies the effect of both the additional sales volumes achieved by the generic parties and the effect of the lower prices on Lundbeck. He concludes that the springboard period for volume concluded on 31 July 2013 (ie about a six month head start). He concludes that the springboard period on LEXAPRO prices ended on 30 September 2015 and on ESIPRAM prices ended on 31 March 2015. … Fourthly, for the purposes of his analysis, Mr Stone was instructed to assume that: (a) for every unit of escitalopram product sold by the generic parties, Lundbeck DK would have manufactured, and CNS Pharma would have sold, a unit of ESIPRAM of the same strength/pack size; (b) the conduct of the generic parties during the initial period referred to above did not change the size of the escitalopram market. In short, Mr Stone's analysis is based on a 1:1 loss of sales. 377 Lundbeck also noted that: The initial period and the springboard period were established in Mr Stone's evidence when all of the generic parties remained in the proceeding. As Sandoz is now the only remaining party, Lundbeck recognises that a minor adjustment is required to these periods to reflect Sandoz' two week early entry. The effect of this is that Sandoz was able to commence selling ESITALO on 1 December 2012, rather than 1 January 2013. However, Lundbeck contends that the effect of this is negligible. 378 On my conclusions about Sandoz's licence reaching the end of its term on 13 June 2009, this adjustment is unnecessary. If my conclusion is incorrect then as explained below Lundbeck is correct that the effect must be negligible.

Some non-issues 379 An interesting aspect of this matter is Lundbeck's approach to the construction of the hypothetical past. I considered this issue in a different context in Sigma Pharmaceuticals (Australia) Pty Ltd v Wyeth [2018] FCA 1556. 380 In Sigma generic companies had been subject to interlocutory injunctions restraining them from entering the market in competition with the originator, Wyeth. As the interlocutory injunctions had been wrongly granted the generic companies claimed compensation pursuant to Wyeth's undertakings as to damages which it had given as the price of the grant of the interlocutory injunctions. As a result, it was necessary to evaluate what the market would have been in 2009 and 2010 had the generic companies been permitted to enter the market (the past hypothetical) when in fact the generic companies had not been able to enter the market for another two years due to the interlocutory and then final injunctions which were set aside on appeal. The generic companies retained an econometric expert to model what prices they would have sold their products at and what their respective market shares would have been had the generics been permitted to enter the market in 2009. 381 In the present case, in contrast, the generic companies, including Sandoz, entered the market in June 2009 and remained on the market throughout the extended term of the patent (until 9 December 2012). Because of the extension of the term of the 144 patent and the operation of s 79 of the Patents Act, the conduct of Sandoz in selling its products between 15 June 2009 and 9 December 2012 infringed the 144 patent for which Sandoz is liable in damages. In short, Sandoz entered the market on 15 June 2009 and thereby infringed the 144 patent. In order to avoid infringement, the earliest date on which Sandoz could enter the market (given that its licence reached the end of its term on 13 June 2009) was 13 December 2012. However, all of the generic companies entered the market on the basis that their products had been listed on the PBS and it is an agreed fact that the earliest date on which any generic company, including Sandoz, could have obtained PBS listing is 1 January 2013. 382 Lundbeck's approach to loss involved the simple step of a time shift. The market shares those generic entrants achieved are known facts. This includes the market shares which Sandoz (and all participants in the market) achieved from time to time and the prices at which Sandoz (and all participants in the market) sold their products. Lundbeck's approach shifted in time the generic market shares achieved at its expense, including the shares of Sandoz, so that their entry to the market commenced on 1 January 2013 rather than 15 June 2009. By this means, instead of requiring econometric modelling to ascertain prices and market shares Lundbeck could rely on the prices and market shares actually involved in generic entry into the market. 383 It may be apparent from this that it was far easier for Lundbeck to prove the value of its loss on the basis of the generics having entered the market than it was for the generics in Sigma to prove the value of their loss on the basis that they were wrongfully held out of the market. 384 To return to the immediate issue, Sandoz made a series of submissions seeking to challenge this approach but these submissions are misconceived. 385 First, Sandoz stressed that the market changed between 2009 and 2012. In particular, pharmacies obtained greater bargaining power relative to pharmaceutical companies, with fewer non-aligned pharmacies (that is, pharmacies not aligned to a particular generic supplier as its first-line supplier). Further, Sandoz had more pharmacies contracting with it as a first or second-line supplier in 2012 compared to 2009. Finally, because on its case Sandoz would have obtained a two week early entry to the market in 2012 it would have achieved better market penetration than other generic companies. 386 There are problems with these submissions. 387 Assume the market did change as Sandoz contended between 2009 and 2012. The fact is that it changed for all participants in the market and it is reasonable to infer that all participants would have conducted themselves accordingly to maximise their competing positions. As such, nothing in the fact of an overall structural change to the industry as a whole would support an inference of different market shares being achieved from those achieved in 2009. To support such an inference some particular change (such as the removal from the market of a generic competitor) would be relevant, but not an overall change affecting all participants in the market in the same way. 388 Sandoz also appeared not to appreciate that if it is the fact that it would have achieved a greater market share from 1 January 2013 than it achieved in 2009 because of its greater number of aligned pharmacies and two weeks early entry to the market then Lundbeck's approach of transposing the market shares Sandoz did achieve from 2009 to 2012 to 1 January 2013 is of benefit to Sandoz. It operates to reduce Lundbeck's loss because it treats Sandoz as being in no better positon from 1 January 2013 than it was in from 15 June 2009. 389 Further, for the reasons already given, Sandoz was not entitled to a two week early entry into the market in 2012 because it did not have a licence to do so. Accordingly, submissions to that effect are immaterial. 390 Second, Sandoz stressed its pricing and discounting strategy referring to evidence of discounts it would have offered from December 2012 based on the discounts it did offer from June 2009. Again, Sandoz appeared not to have appreciated that it is inherent within Lundbeck's approach of transposing the actual market from June 2009 to January 2013 that Sandoz is taken to have done in 2013 exactly what it did in 2009 including by offering the discounts it did. And again, for Sandoz to submit, as it did, that it is "is unclear what share of the market in 2013 Sandoz would have captured with such a discounting strategy given the differences between the market in 2009 and 2013 is an odd submission for it to make because, if accepted, it would involve an increase in Sandoz's market share and thus an increase in Lundbeck's claims for damages. Equally counter-intuitive is Sandoz's related submission that: Mr Turner gave evidence that it was his expectation that CNS Pharma would have discounted Esipram in Scenario 2 from 1 January 2013 at the rate that it actually did in the real world from June 2009. The Applicants chose to adduce no evidence in response to this evidence, and did not cross-examine Mr Turner on this evidence. It is to be assumed that the Applicants' discounting strategy in Scenario 2 would have been the same as or similar to that which Mr Turner has described (as best he can). 391 As Lundbeck noted in its closing oral submissions, this submission: … is another example of our learned friends putting forward the very point we put forward, namely, we would have done - I'm not quite sure what the issue is between us. They say they would have done what they did; they say we would have done what we did, and that's why it's a good blue print for working out the springboard period. 392 That this is so is demonstrated by Sandoz's subsequent submission that: … from January 2013 and thereafter, Sandoz's discount to AP2P would have been the discount that it actually offered one month after the generics began selling escitalopram products in Australia in the real world. Sandoz's generic competitors were likely to have offered the same discounts in 1 January 2013 that they offered when they launched in the real world in 2009 and CNS Pharma is likely to have discounted from 1 January 2013 at the same rate that it discounted in the real world in June 2009. 393 Third, Sandoz submitted this: In the counterfactual, from at least 1 January 2013 onwards when all the generics had escitalopram products on the market, the substitution rates are likely to have been different to those in the real world in mid-2009 and there could have been more substitutions of other generic brands, including Esitalo, in lieu of sales of Esipram in that period. This does not support the unproved assumption adopted by the Applicants that the counterfactual market shares in 2013 would have been the same as the market shares achieved in the real world in June 2009. 394 Again, coming from Sandoz, the submission seems counter-intuitive. It also fails to confront the fact that nothing suggests that any individual generic entrant to the market would have conducted itself differently in 2013 from how it in fact conducted itself in 2009. Sandoz's approach also overlooks the fact that no hypothetical construct can be known with certainty. As the Full Court said in Bayer, a hypothetical construct, be it past or future, is necessarily the outcome of an evaluation of the probabilities and possibilities. In Sigma at [301], I put the matter this way: The so-called "counter-factual" approach is merely a tool to assess whether the operation of the interlocutory injunctions has caused loss. 395 By analogy, in a claim for damages for patent infringement the past hypothetical construct is a tool to assess loss. The construct will always be the result of an overall evaluation involving greater and lesser degrees of uncertainty, estimation and even speculation. 396 In the present case, we have an extraordinarily useful tool which, in my view, tends to minimise the uncertainties. We have the fact that the same generic entrants to the market in 2009 are those who remained in the market in 2013. To identify what losses were suffered as a result of Sandoz (along with other generic companies against which the claims for damages have been settled) entering the market on 15 June 2009, when it could not do so without infringing the 144 patent until 1 January 2013, transposing what occurred from 15 June 2009 to 1 January 2013 is nothing more than good sense. It is the best possible evidence of what would have occurred but for the infringements of the 144 patent. 397 Fourth, Sandoz submitted this: The two scenarios which the Applicants contend for and on which Mr Stone bases his calculations ('Scenario A' and 'Scenario B') assume that Apotex, Alphapharm, Aspen and Sandoz all entered the escitalopram market on 1 January 2013. However, even on the Applicants' interpretation of the Sandoz licence, Sandoz was licensed to enter the market earlier than this, namely on 1 December 2012. On no view, therefore, can either of the Applicants' counterfactual Scenarios A or B be correct. Scenarios A and B are irrelevant, and by not proving their loss under Scenario 2, the Applicants like the Applicant in Aristocrat have quantified their loss on a false basis. 398 For the reasons already given I reject this submission. Sandoz did not have a licence to enter the market on 1 December 2012. In any event, even if this is incorrect, the case bears no resemblance to Aristocrat. Sandoz itself noted that Lundbeck had adduced evidence that the two week period would make no material difference to the price reductions. As Sandoz said: The only evidence that the Applicants called in relation to Scenario 2, was Ms Weber's lay evidence in Weber 3 as to the price reductions that would have occurred in the springboard period from January 2013 to December 2015. In Weber 3, Ms Weber contends that the consequence of Sandoz entering the market on 1 December 2012, was that the voluntary reductions would be brought forward by one month and that there would be no material change to the price reductions identified in Weber 1 at [28]-[31], as corrected in Weber 3 at [6]-[16]. 399 As Lundbeck said, Sharon Weber: …was Senior Market Access Manager employed by Lundbeck AU. She is a Pharmacist with a graduate qualification in Pharmacoeconomics. Her roles with Lundbeck AU were focused on pharmaceutical pricing in Australia, including the effects of government pricing schemes such as the PBS. Ms Weber gave evidence about the price of LEXAPRO and ESIPRAM between June 2009 and June 2014, including the effect of PBS price disclosure, statutory price reductions, voluntary price reductions and (in the case of LEXAPRO) brand price premium increases. Ms Weber also calculates what the prices for LEXAPRO and ESIPRAM would have been: (a) between June 2009 and December 2012, had no generic escitalopram medicines been launched during that period; and (b) between January 2013 and December 2015, had generic escitalopram medicines been launched in about January 2013. In Weber 3, Ms Weber also comments on the effect of Sandoz entering the market on 1 December 2012, rather than 1 January 2013. She considered that this would have a negligible impact on her calculations. Ms Weber was not cross-examined and Sandoz did not put forward any evidence challenging Ms Weber's calculations. Ms Weber's evidence therefore stands unchallenged. 400 As Sandoz did not cross-examine Ms Weber, it cannot in submissions challenge her evidence that the two week early entry by Sandoz (if relevant at all, which I consider it cannot be) would be immaterial. 401 Fifth, Sandoz said: In calculating damages in Scenario A and B, Mr Stone was instructed to make the following assumptions as to counterfactual market share. In Scenario A, he was instructed to assume that each of Apotex, Alphapharm, Aspen and Sandoz captured the same proportion of the generic escitalopram market from 1 January 2013 that they in fact captured from June 2009 onwards, and that each of Apotex, Alphapharm, Aspen and Sandoz entered the market in the same order and magnitude from 1 January 2013, as they did from 15 June 2009. In Scenario B, Mr Stone was instructed to assume that Apotex, Alphapharm, Aspen and Sandoz entered the market together on 1 January 2013, and that the market shares for each of Apotex, Alphapharm, Aspen and Sandoz from January 2013 to April 2013 would have been the same as the actual market percentages achieved in October 2009. The Applicants have adduced no lay or expert evidence to prove the assumptions as to counterfactual market share underpinning Scenarios A and B. Those assumptions are a fundamental basis upon the Applicants' losses in Scenarios A and B have been calculated by the expert accountants. The Applicants bear the burden of proving counterfactual market shares on the balance of probabilities and with as much precision as the subject matter reasonably permitted. There can be no suggestion that it was not possible for the Applicants to adduce lay or expert evidence to support their instructions to Mr Stone on these matters. Rather, this is an instance where the plaintiff has simply failed to do so. In Placer (Granny Smith) Pty Limited v Theiss Contractors Pty Limited (2003) 196 ALR 257 at [37]-38], Hayne J pointed to the stricter approach taken by the courts in assessing damages where a plaintiff has not adduced evidence that was apparently available to prove the loss. 402 It should be apparent that I do not accept that Lundbeck has failed to prove the assumptions informing its approach to hypothetical market shares. It has proved the necessary facts in the most reliable manner possible by taking what actually occurred from 15 June 2009 and running the time-line from 1 January 2013 in the two different ways described as Scenario A and Scenario B. For my part, I see no reason to depart from Scenario A which simply transposes the time-line from June 2009 to January 2013. Accordingly, I see no reason to dwell on Scenario B. 403 Sixth, Sandoz made other submissions about mandatory price disclosure requirements and its consequences which it was not at liberty to make because it elected not to cross-examine Ms Weber. 404 Seventh, Sandoz made this submission: The Applicants have adduced some limited evidence to the effect that Lundbeck DN would have had the manufacturing capacity to meet increased demand for Lexapro and Esipram in the Australian market in the counterfactual. Evidence of the manufacturing capacity of Lundbeck DN was provided by Mr Carsten Andersen, Vice President of Group Finance Lundbeck DN. There is no explanation in his affidavit as to why he would know anything related to the Lundbeck DN's manufacturing capacity as a result of his performance of his duties in finance at Lundbeck DN since March 2010. His evidence amounts to little more than assertion, without supporting business records or reasoning, that Lundbeck DN had sufficient manufacturing capacity during 2009 to 2015 to meet increased Australian market demand for approximately 1.2 million additional boxes of escitalopram 10mg x 28 tablets, and approximately 1.2 million additional boxes of escitalopram 10mg x 28 tablets. The relevant parts of Mr Andersen's affidavit were conclusory in nature and admitted into evidence over Sandoz's objection; the form of the evidence did not place Sandoz in a position where it could test the evidence by cross-examination. Having regard to the fact that the Applicants could have gone about proving capacity properly, the evidence should be given no or little weight. 405 The submission is without merit. Mr Carsten was responsible for evaluating the sales that would have been made by Lundbeck AU and CNS Pharma for all of Lundbeck's internal and external accounting purposes. He had to know that there was sufficient manufacturing capacity to enable sales targets to be met. To suggest otherwise is a nonsense. The idea that Lundbeck, an international pharmaceutical company which had sold its products in the Australian market for many years, could not manufacture the drugs to meet the market demand that would have existed but for Sandoz's infringement of the 144 patent fails to recognise that Lundbeck and its subsidiaries were involved in a serious business endeavour on an international basis. There cannot be any real doubt that Lundbeck would have been able to meet the increasing demand in the Australian market just as it had been doing for years before Sandoz's infringing entry into the market. 406 I can now move to what I consider to be the real issues about the damages claims.

Substitution 407 Lundbeck's claims are based on the proposition that every Sandoz escitalopram product sold represents a lost sale of its cheapest escitalopram product, being the product sold by CNS Pharma known as Esipram. 408 Lundbeck submitted that this approach was justified on the evidence and, indeed, conservative because the claims do not include any lost sales of the more expensive escitalopram product, Lexapro, sold by Lundbeck AU. The claims in respect of Lexapro are confined to the reduced sales prices which were charged for Lexapro due to the generic entry into the market. 409 Lundbeck said: In the April hearing, the Court heard evidence from Lundbeck's prescriber witnesses, Professor John Tiller (a psychiatrist) and Associate Professor Grant Blashki (a general practitioner). …They gave evidence about the circumstances in which, and how, they prescribe escitalopram. The key point which emerges from their evidence is that the launch of generic escitalopram did not alter their prescribing practices. They continued to prescribe escitalopram in accordance with the clinical criteria they used throughout the relevant period. They were cross-examined briefly on 1 May 2018. They were both careful and responsive witnesses. Their evidence should be accepted without reservation. In cross-examination, the generic parties focussed their attention on the cost of escitalopram. However, neither Professor Tiller nor Associate Professor Blashki were moved from their evidence that, during the relevant period, cost was a very infrequent and very marginal consideration in their prescribing practices for anti-depressants. In short, they did not prescribe more escitalopram because cheaper generic medicines became available. The generic parties did not put forward any evidence in response to the evidence of Professor Tiller and Associate Professor Blashki. 410 Sandoz noted that although the accountant it called, Mr Samuel, adopted the same approach to substitution of Lundbeck's products for Sandoz's products (that is, 1:1) and Sandoz did not challenge this approach "in the sense of suggesting that it should not be applied in the course of calculating any losses", it submitted that there was nevertheless a risk that not all sales of its products represent a lost sale of Lundbeck's products, supporting the imposition of a discount for risk on the overall claims. 411 In support of this proposition Sandoz made these points: (1) The antidepressant market from June 2009 onwards, was "crowded" with other antidepressant products in competition with Esipram, so that doctors had other choices available to them; in particular other selective serotonin reuptake inhibitors, serotonin and noradrenaline reuptake inhibitors, tricyclic antidepressants, monoamine oxidase inhibitors, and tetracyclic antidepressants. (12) When prescribing an antidepressant in the treatment of depression or anxiety disorders, the class of antidepressant is first chosen based on a number of factors relating to the clinical presentation and history of the patient, it being a matter of trial and error to figure out which of the SSRIs and SNRIs would suit the individual patient. (13) The cost sensitive consumer would not have been attracted to purchase Esipram where a cheaper SSRI was available - they would have asked their doctors to prescribe an alternative SSRI molecule. (14) Professor Tiller and Associate Professor Blashki: …did consider cost when prescribing a class or brand of medicine in circumstances where the patient raises cost as a concern, when prescribing an antidepressant that is not subsidised under the PBS, when the patient is in a severe financial situation and when no generic version of a particular medicine is available. Their evidence that cost is typically not a 'significant factor' in the decision which antidepressant to prescribe is to be understood in the context that in the main the molecules of antidepressants they have prescribed have had alternative generic brands available. For cost-sensitive consumers, Professor Blashki and Prof Tiller may select instead of escitalopram another SSRI molecule that was the subject of generic competition (generic being sold at lower than the benchmark premium price were available). 412 Professor Tiller's evidence was that if a patient had a concern about cost he would take that into consideration but: …in reality because of the Pharmaceutical Benefits Scheme and its availability for most antidepressants, the financial impost is relatively minor even for those in straightened circumstances. So it is not normally an issue at all. 413 Accordingly, cost "would not stop you treating the patient in the most appropriate way you can". 414 He reiterated that: Financial pressures with a PBS substance would be extraordinarily uncommon in my experience. It is not in day-to-day life a factor of any material consequence… People who are pensioners and with handicapped families and difficult financial situations, but the actual cost of these medicines is relatively modest and for some it is with the safety net it is effectively little or nothing or for others it's little more than a cost of a cup of coffee and most in the course of a month can manage that. So in real life it's not an issue. 415 Professor Tiller accepted that in his experience the antidepressants he prescribed had generic equivalents available. This exchange then occurred: Would you not accept that there may be a difference to the patient in terms of cost implications between a medicine which is on the PBS with no generic competition to a medicine which is on the PBS which has generic competition? In practice, that's of no consequence. They're both PBS subsidised and - how do I put it - with these list of factors for the choice of medicine, there has not been a weighting provided to each of those reasons. They might be read as if they are of equal weight. They are absolutely not. It is the clinical status of the patient and their history and those things which matter. And the cost and economics absolutely would be seen as a most trivial part of that process in regular clinical practice. That is for PBS-subsidised agents. … I have had many meetings with many psychiatrists all around the country in quite wide area - different areas of practice. And in that, where we've had robust discussions about factors that relate to choice of medicines, adherence, treatment, and the like, the cost has not been an issue that has been raised by others. 416 Associate Professor Blashki's evidence was that: Cost is a minor consideration as a GP, especially when weighed against the enormous impact of a mental health condition on someone. And in my 25 years as a GP I can't recall a single time that it has come up… It's just not an issue that I've discussed with people, the cost of an antidepressant, to the best of my memory. 417 Associate Professor Blashki accepted that the products he prescribed had generic equivalents available and that, if this was not the case (which he had never experienced), then it would be reasonable to consider cost, but the fact remained that: … cost is a very minor consideration at all with the prescribing of anti-depressants with my patients, even in St Kilda, which is pretty low socioeconomic. It is not a conversation we have. And the reason for that, if I may, is that the burden of mental health issues for people is non-trivial and that anything that can help them recover quickly is - dramatically offsets any small economic cost. 418 Associate Professor Blashki agreed that if a patient raised cost he would consider it but said: This assumes an unreal world of this - of this consideration as a GP. This is not a conversation I've had in my mind or with my colleagues ever. The main drivers of my choice of medication are familiarity with the medicine, based on the principle which I actually teach to my students, as well, which is to get to know a few brands well. But I do accept that theoretical position, and the common sense answer as to what that might be, which that I would take it into account. 419 The overall effect of this evidence is this. Doctors prescribe the antidepressant that they think will be clinically appropriate for the particular patient. In practice, provided the antidepressant is PBS listed (which Lexapro was), cost is not an issue. This is so whether or not there is a generic equivalent product. 420 For these reasons, Sandoz's submissions are over-stated. If escitalopram is not or proves not to be the right antidepressant for the patient, it will not be prescribed or will cease being prescribed. Such a patient, accordingly, would either never be prescribed Sandoz's escitalopram products or will have ceased being prescribed Sandoz's escitalopram products. As a result, the starting point must be that every sale of Sandoz's escitalopram products is a lost sale of a Lundbeck escitalopram product. 421 The reason I say that in this case 1:1 substitution is the starting point is that, as we know, Sandoz's escitalopram products were not the only generic equivalent escitalopram products on the market. As Sandoz identified, from mid-2009 there were multiple generic equivalent escitalopram products available. The fact that Lundbeck settled its claims against these generics does not mean that their products may be disregarded. While I would accept that there is a very high probability (approaching a certainty) that every sale of any one of these generic equivalent escitalopram products must represent a lost sale of Lundbeck's escitalopram products, I do not accept that every sale of Sandoz's escitalopram products involves a near certainty of representing a lost sale of Lundbeck's escitalopram products. In a crowded market of multiple generic equivalent escitalopram products from mid-June 2009 onwards, some sales of Sandoz's escitalopram products must represent lost sales not of Lundbeck's escitalopram products but of other generics' escitalopram products. 422 The fact that those other generics were in the market cannot be ignored. Sandoz is not responsible for the conduct of those other generics and cannot be held liable for it. This indicates that Sandoz must be right, albeit not for the reasons it gave, that there should be a material discount to account for the probability that some Sandoz sales otherwise would have been sales of other generics' escitalopram products rather than Lundbeck's products. The materiality of this necessary adjustment is to be assessed having regard to the other factor which I noted above, that Lundbeck has approached the matter on the basis that every lost sale is a sale of the cheapest product offered by CNS Pharma (Esipram) rather than the more expensive product offered by Lundbeck AU (Lexapro). There is a very high likelihood that some sales of Sandoz's escitalopram products would represent lost sales of Lexapro rather than of Esipram. Professor Tiller's evidence explained why this would be so. He said this: …you initiate the treatment and after a few weeks you - or may rapidly find out if it's tolerated or not, but after a few weeks determine whether it is effective or not. But then a really important element for the patient, I believe, is consistency of brand, packet, appearance of pills and so forth, because the patient is then familiar with their treatment and also becomes aware in the occasion when there might be a prescribing or dispensing blunder, they will say something has changed, so that consistency is also very important part, in my view, for… patients, treatment and maintaining that. 423 As a result, it is reasonable to infer that some patients who had been taking Lexapro would have been likely to keep taking Lexapro but instead were substituted to Sandoz's escitalopram products. 424 In my view, while it is preferable as a matter of principle to make one overall adjustment for risk at the end of the damages assessment process, this issue is likely to be the most material factor to take into account in deciding the size of the adjustment which should be made. There is little evidence to assist in resolving what size of adjustment would be appropriate to account for the fact that some sales of Sandoz's escitalopram products must represent lost sales not of Lundbeck's escitalopram products but of other generics' escitalopram products. There is some evidence that generic substitution rates were increasing over time. Whilst Lundbeck characterised the evidence of Jason Turner of Sandoz about increasing generic substitution rates as unreliable, noting that "he could not name a single pharmacist who had informed him of this", and had been Sandoz's Chief Financial Officer in 2009 (only subsequently moving into a business sales management role), I do not have any concern accepting Mr Turner's evidence. Even in his role as Chief Financial Officer he must have had a good sense about the market within which Sandoz was operating which was all about generic substitution. It also stands to reason that over time pharmacists and patients will become more familiar with the process and potential benefits of generic substitution. 425 The result is that, doing the best I can on the evidence, I consider that a material discount from the assumed substitution rate of 1:1 is required. Resolving remaining doubts in Lundbeck's favour to reflect the general principle that a liberal approach to damages is required and having regard to the fact that Lundbeck's approach excludes any claims for lost sales of Lexapro, I consider that a deduction of 25% is required to account for the fact that it is likely that some sales of Sandoz's escitalopram products must represent lost sales not of Lundbeck's escitalopram products but of other generics' escitalopram products.

The claims for infringement 426 It is convenient here to summarise the claims made for damages in relation to Sandoz's infringement of the 144 patent. 427 As noted, at all material times Lundbeck A/S was (and is) the proprietor of the 144 patent. Lundbeck AU, on my conclusions, was (and is) the exclusive licensee of the 144 patent which supplies Lexapro in Australia. Sandoz infringed the 144 patent. Lundbeck A/S and Lundbeck AU each claim damages for the infringement of the 144 patent. CNS Pharma, on my conclusions, has been authorised by Lundbeck AU to exploit the invention claimed in the 144 patent which it does by supplying Esipram in Australia. CNS Pharma, being neither the patentee nor the exclusive licensee of the 144 patent, cannot and does not claim damages for patent infringement. This fact, it might be thought, would be the end of the matter insofar as lost sales of Esipram are concerned. Not so, however. 428 Lundbeck A/S's claim for damages as the patentee of the 144 patent includes the money it would have received from CNS Pharma on the sales of Esipram which were lost to CNS Pharma as a result of Sandoz's infringements. Lundbeck A/S would have received this money because Lundbeck AU and CNS Pharma are wholly owned subsidiaries of Lundbeck A/S and are subject to a group-wide transfer pricing policy arrangement under which there was an annual adjustment made having regard to all sales by Lundbeck AU and CNS Pharma (the vast majority but not the totality of which were sales of Lexapro and Esipram respectively) to ensure that those companies each achieved a return on sales (or ROS) of approximately 3%. The adjustment would be effected by way of either a credit note (if the unadjusted return on sales was less than the 3% target) or an invoice (if the unadjusted return on sales was greater than the 3% target). 429 There is a debate between the parties, and the accounting experts, about whether Lundbeck A/S can sustain this claim. If it can be sustained, there is a further debate about the claim period for CNS Pharma.

19.1 The transfer pricing policy 430 Carsten Andersen, Lundbeck A/S's Vice President, Group Finance, between 2014 and mid-2018 gave evidence about the transfer pricing policy arrangements between the companies in the group. He explained that Lundbeck A/S sold Lexapro to Lundbeck AU and Esipram to CNS Pharma at a transfer price. At the end of each year Lundbeck A/S would consider whether to make an adjustment so that each subsidiary achieved a target return on sales of approximately 3%. In his cross-examination Mr Andersen confirmed this process. As he explained the policy is continuously applied: …throughout the year, you're setting a transfer price to align itself with the budget. So if everything goes smoothly according to the budget, the transfer price will yield the allocation of profits that you want to do to adhere to the policy. If there are any changes throughout the year that you capture to realign the baseline, you will reach that result also. If for some reason things happen maybe late in the year or you determine for resource purposes or timing purposes not to adjust it, but see what happens, and at year end you do require an adjustment up or down to sort of like get back in the range, that is also the transfer price. It's just - it's just a matter of how it's done practically and when it's done from a timing perspective. 431 I consider that there is no doubt the policy existed at all material times and was consistently applied as between Lundbeck A/S, Lundbeck AU and CNS Pharma in the manner Mr Andersen described.

19.2 The accounting experts' views about the transfer pricing adjustment 432 In their joint report Mr Stone and Mr Samuel agreed that but for Sandoz's infringements the year-end transfer pricing adjustment would have been different as a result of Lundbeck AU and CNS Pharma each making additional sales of their products. They also agreed that: The quantum of any additional hypothetical Year End TPA can be determined by applying the transfer price agreements that were in place at the relevant times. Mr Stone's calculations are based on those agreements. The mechanism he has used is that which was used to generate the actual Year End TPAs during the relevant period. 433 Other agreements included these matters: Any such calculation of quantum assumes (and Mr Stone's calculations of the hypothetical Year End TPA assume) that Lundbeck Australia and CNS would not have spent any of the additional profits they would have generated in the hypothetical scenarios on matters unrelated to Esipram and Lexapro. No evidence has been provided in this proceeding, to the best of the knowledge of the Experts, as to whether and if so to what extent any of the extra profit (which exceeds $65 million in all scenarios) would have been spent on matters unrelated to Esipram and Lexapro. The Experts have not attempted to speculate as to what any such expenditure might be. If additional expenditure would have been incurred which had no impact on profitability other than the cost itself, it would have had an effect of reducing the adjustment to the Year End TPA. The Actual Year End TPAs were not allocated by product. … The specific actual policies in place, through the use of a targeted return on sales (measured on an overall basis), give rise to one adjustment, and not a series of adjustments for each product or product group. … The Year End TPA is a consequence of all the transactions entered into during any financial year. It is not an input to the determination of the profit on any one product. 434 Mr Samuel agreed with Mr Stone's arithmetic in calculating the hypothetical annual adjustments that would have been paid to Lundbeck A/S, but said that the claim for damages could not be sustained because: a) each sale is a separate infringement or contravention of the ACL. No Year End TPA arises as a consequence of one sale. Mr Samuel further notes that any award made prior to the end of the year in which infringement would have been made would not take into account any change to the Year End TPA. For example, if an infringement in April 2010 led to an award for damages in November 2010, there would be no adjustment to the award to account for the Year End TPA. Mr Stone's adjustment arises only as a consequence of the timing of the award; b) the Year End TPA is a consequence of a myriad of factors relating to all sales of all products, all expenses incurred and the transfer prices determined for all products for each year, as a matter of discretion by H Lundbeck; c) it is not possible to say with certainty how any extra funds would have been spent by Lundbeck Australia. Mr Samuel notes that this is a different question to whether expenditure was necessarily incurred for the purpose of calculating the damages arising from the infringing sales of Esipram and Lexapro; d) the loss of profit suffered by Lundbeck Australia and/or CNS arises at the date of sale. In other words, the Year End TPA is not a cost necessarily incurred in order to make a sale of Esipram or Lexapro. It is instead an adjustment that arises as a consequence of the profit made on the sale of those products (as well as all other profits, losses and expenses) having been made during the course of the year. In this regard, it is more akin to a distribution of profit, in that it only arises once the total profit for Lundbeck Australia and CNS combined for the year has been determined or can be reasonably estimated. For the avoidance of doubt, Mr Samuel is not suggesting here that the Year End TPA is a dividend, only that the Year End TPA is a consequence of the profit being made on Esipram and Lexapro - it is not an input to the determination of that profit; e) the Year End TPA is not allocable by product. Again, the Year End TPA is a consequence of the profit being made on Esipram and Lexapro (as well as all other products) - it is not an input to the determination of that profit. It follows that the Year End TPA is attributable to all transactions the companies engaged in. Mr Samuel agrees with the mathematics of Mr Stone's hypothetical Year End TPA, in total. However, it cannot be said that the whole of the adjustment is attributable to Esipram and Lexapro, as the increase in profits would result in different proportions of the total Year End TPA being attributable to each product, see example at Appendix B. and the consequence of attributing a hypothetical Year End TPA in these proceedings to the lost sales of Esipram and Lexapro is to distort the gross profit margins attributable to Esipram and Lexapro. 435 Mr Stone and Mr Samuel discussed these issues in their oral evidence which was given in a concurrent session. 436 In respect of his proposition (a), Mr Samuel agreed that if Lundbeck AU and CNS Pharma had sold additional products then Lundbeck A/S would have received more money by way of the annual TPA (transfer pricing adjustment) than it in fact received during the periods affected by Sandoz's infringements of the 144 patent. He agreed also that, on the assumption adopted as to additional sales that would have been made but for Sandoz's infringements, Mr Stone's calculations of that increased amount were correct. This exchange confirmed this position: MR BANNON: Yes. Speaking as an accountant, if you're asked the question, assuming application of the policy and assuming the Stone figures, would Lundbeck have ended up with more money in its pocket if the infringements hadn't occurred, you say yes, it would have. MR SAMUEL: Ultimately it would have, yes. 437 Mr Samuel explained that: … my issue is the TPA is not then put to the damage. The TPA is a consequence of the profit made on the sale of the products, so there's an issue - my issue is whether there should be an adjustment at all. 438 The exchange continued in these terms: MR BANNON: Yes. I think the point you make in your - the report and in the joint report is you query or aren't satisfied that that's - the approach I suggested, are you financially better off, you are not satisfied that's the correct approach to apply in working out damage for infringing sales; is that right? MR SAMUEL: That's correct. MR BANNON: But if it might - I will call it simplistic approach. Is Denmark going to be better off money-wise - if that's the correct approach, then you would agree that, again assuming the Owain Stone, the Stone calculations, you would agree they would be better off by that amount. MR SAMUEL: Yes. If the court concludes that's the way to go, I'm in agreement with the mathematics. 439 Mr Samuel explained his position in greater detail in this way: … the TPA is not an input to the profit made as a consequence of the infringement or the profit lost as a consequence of the infringement. It is a consequence itself of the profit made as a result of the infringement. So what you're doing is putting the TPA as an input to the calculation of the lost profit, and I'm saying you have to calculate the lost profit. That helps you determine, together with thousands of other transactions, what a TPA would be. … …this is all about timing only. We're only having this discussion because the award postdates actual year end TPAs. 440 Mr Samuel also said this, for example: The loss is suffered in CNS. CNS receives the award and they do a year-end TPA, and no doubt transfer some money to Denmark. So Denmark ends up back where it should be in the normal course of business. 441 CNS Pharma, of course, was neither the patentee nor the exclusive licensee of the 144 patent. It makes a separate claim for misleading and deceptive conduct but, insofar as the 144 patent is concerned, it cannot claim for its lost sales as a result of Sandoz's infringements. Lundbeck A/S, however, claims in its capacity as the patentee for the loss it has suffered as a result of the lost sales of both Lundbeck AU, the exclusive licensee, and CNS Pharma, authorised by Lundbeck AU to exploit the invention in Australia. Accordingly, insofar as compensation to Lundbeck A/S is concerned from Sandoz's infringements of the 144 patent, it is not the case that the loss suffered by CNS Pharma and Lundbeck A/S will not end up "back where it should be in the normal course of business" (leaving aside an award of damages to Lundbeck A/S for loss in its capacity as patentee). 442 This evidence was then given: MR BANNON: I understand. But you understand that you were made aware that - if CNS only suffers a loss, then that may not be recoverable at all by any Lundbeck company. That's an argument Sandoz wants to put forward. MR SAMUEL: Yes. Well, this dovetails into a legal issue I can't help with, which is it seems to me that if you put the TPA into the calculation of damages, you're converting a non-compensable loss into a compensable one, but that crosses the line of what I can assist with. … MR BANNON: But if, absent the infringements, Lundbeck Denmark would have received the amount of a TPA, extra TPA, but CNS can't get the claim for some legal reason, then that's a loss of Lundbeck Denmark, isn't it? MR SAMUEL: Lundbeck will have less than it would otherwise have had, yes. … I don't think it should be feeding into the calculation of loss at all. … It's not actually distribution of profit, but it's an internal transfer between related parties. It's not an input to the calculation of the loss of profit arising from the infringement. MR BANNON: Aren't you really jumping into a legal test of causation, rather than confining yourself to an accountant's approach of whether a company is better or worse off financially? MR SAMUEL: No, I don't think so. What I'm trying to work out is what the financial consequences of the infringement. Certainly, I have to - that means there's a causal link. Somewhere between you and me there's a bridge, law on one side and accounting on the other. My accounting has to follow whatever that causal link is, but I can't calculate the profit by inputting something from the TPA; I can only calculate the profit by looking at the costs incurred in generating an incremental sale. MR BANNON: But what you can do as an accountant is say would Lundbeck Denmark have been better off financially if these infringements had not been committed. MR SAMUEL: Yes. And I've agreed with that. MR BANNON: And, answering that question, it would be perfectly appropriate to take into account the TPA, wouldn't it? MR SAMUEL: I agree they would have been better off. 443 I agree with Mr Samuel's first observation. I consider that the issue with which he is grappling is a legal issue rather than an accounting issue. It is an issue concerned with the conceptualisation of Lundbeck A/S's loss in its capacity as patentee and the principles which govern the award of damages for such loss. 444 In respect of his proposition (b), Mr Samuel agreed that as between the accountants there was no dispute about lost sales and no dispute that the cost of sales had to be accounted for, nor any dispute about the transfer prices being determined on the basis of all products for each year. Mr Samuel said "I actually think Mr Stone and I are agreed pretty much on all the assumed facts". Mr Samuel accepted that the transfer pricing policy had to be applied consistently from year to year, noting that he had seen similar arrangements in other international pharmaceutical companies. He explained his concern in these terms: So there's multiple products in here, not just Esipram and Lexapro. So if they were amending it, as you suggested earlier, on a quarterly or semi-annual basis, you would be amending the transfer prices for other products, not just Esipram and Lexapro. MR BANNON: In other words, how you shift the prices around to achieve the ultimate overall result may involve some discretionary element. MR SAMUEL: Yes. MR BANNON: But the overall result will always look the same - well, within the guidelines, at least, as you understand it. MR SAMUEL: Yes. The question is what is it attributable to? 445 In respect of his proposition (c), Mr Samuel said: I can't say what extra expenses would have been incurred, but we're in a scenario here where we have a sea change to the volumes of the business as between the actual and the hypothetical world. In 2012 the hypothetical units sold at 2.3 times higher than the actual units that were sold. So we have a whole of organisation consequence going on here. Any dollar, for whatever reason, spent by Australia or CNS would affect the TPA is the point I'm making. … Let me clarify. I apologise. I'm not talking here about expenses which are a consequence of the infringement. I'm talking about any other expenditure unrelated to Esipram and Lexapro MR BANNON: Why would that have any effect on TPA, if it's money which should go back to Denmark under the TPA? MR SAMUEL: Well, because if Australia spends $3 million on anything else, even if it chooses send its executives on a jolly to Tahiti, that would be money that comes out of Australia's pocket and it would be fed into the year end TPA adjustment. … What would be your basis for thinking that Australia could dip their hand into the fund which should be going back to Denmark under the TPA to use for anything else, a trip to Fiji or whatever other examples you provided? MR SAMUEL: Well, for example, there's a letter appended to Mr Murray's affidavit from the managing director which talks about lower investment as a consequence of a lower sales base, so that suggests to me that if they had had more money, they would have spent more money. MR BANNON: But have you also seen documents to suggest that there are rechargeable expenses to head office? MR SAMUEL: I'm aware that there are expenses described as defrayed. I haven't seen documents which tell me what that is or what it's for or how it came about. 446 In respect of his proposition (d), Mr Samuel said: … same point I made before. The TPA is not feeding into the profit lost as a consequence of the distribution - consequence of infringement or contravention. It is itself a consequence of the profit that is made on the sale of those products, as well as thousands of other transactions. MR BANNON: But the prices are only provisional, as you understand, it subject to TPA adjustments. MR SAMUEL: Yes. No. I understand that. MR BANNON: Well, doesn't that answer your concern in (d)? MR SAMUEL: Well, no, because what we end up with here then is, let's take the hypothetical - and let's just take your earlier suggestion that they might have made quarterly or semi-annual adjustments. It means the TPAs would have been - sorry - the hypothetical transfer prices for all products would have been amended during the course of the year. So we don't know how much of this TPA is attributable to these infringements. We just don't know. It's an agreed point. The TPA is not allocable by product. Now, that applies whether you're looking at the actual TPA or whether you're looking at the hypothetical TPA. It's an agreed point. We cannot allocate it by product. So to say that this certain amount is attributable to the infringements is wrong. 447 Mr Stone then gave evidence about his views. He confirmed that if Lundbeck AU and CNS Pharma had made the sales that Sandoz had made (referred to as incremental or additional or lost sales) then Lundbeck A/S would have been paid more money by the end of year TPAs for each relevant year. He explained: …I would agree that the year-end TPA is a consequence of a profit being made on all profits and it took, to quote Mr Samuel, thousands of other transactions. My calculation is the calculation of the effect of the change. It's not the original calculation. It is simply saying that calculation would have been different in a hypothetical world, only as a result of a change in the revenues from the two relevant products. And, therefore, the change in the TPA must be a function of the change in the sale of those two products. I've got no information to suggest that they would have taken that money and used it for any other purpose… So, from my point of view, it's correct to say that the original TPA is a function of a myriad of other transactions, but the change in the TPA is not a function of a myriad of other transactions; it's a function of one change, the change in the infringing product sales and margins and costs that go with that. So, from my perspective, I'm struggling to understand why there's any difficulty in estimating the impact of the change, notwithstanding that the original one is not allocated by product. The difference can only relate to these products, because it's only those products that change. 448 In response to a question I asked, Mr Samuel gave this example of his concern: Let's assume the actual TPA was 100 and involved 1000 transactions. And we're agreed we can't allocate that by product. The company doesn't do it. No one is suggesting it can be done. … Whatever the prices are. They are what they are and they're being set. … But, nevertheless, those prices establish profit margins in CNS and Australia which are being generated each time they make a sale, or would have made a sale absent an infringement. … And those margins would have been set - and we know, from looking at their records, that the margin on Esipram, Lexapro, generally speaking, was in the 50 to 60 per cent range. … … This is before you get to the TPA. … All my point is that - let's say there is - we now add in the extra sales and profits, and so now we have a year-end TPA of 150, only now it's due to 1200 transactions instead of 1000, we still have the same problem; we don't know which products that relates to, and yet what we're doing here is allocating the entire increment to the infringing sales. … The entire increment. So if the TPA has gone from 100 to 150, Mr Stone's calculation, which I agree mathematically would take the 50 and so that's attributable to the infringing sales. And I'm saying, well, we don't know that. And, in fact, the consequence of it is that you're now assuming every infringing sale was sold at a margin of three per cent. If you assume no other overheads, you would be saying that every sale was being sold - every infringing sale was sold at a margin of three per cent, where in fact every actual sale was made at a margin of 50 or 60 per cent. 449 Mr Stone responded: I'm bemused by Mr Samuel's explanation, because I'm struggling that if you have increased the TPA by 50 as a result of 200 transactions, then that 50 is caused by the 200 transactions. There is no allocation of that TPA, whether the hypothetical or the actual, to the individual products. The gross margin stays as a - the TPA stays as a single item in the trial balance so it's not allocated to the products. So by his own example, he's saying, well, I can't take the 50 and allocate it to the 200, but the 50 is caused by the 200 under his own example, so I don't understand why it is not caused by the 200. 450 Mr Samuel said: Well, let me respond to that. Here is the problem. That can't be right. That simply can't be right because we only have one adjustment. It's one year-end adjustment, 1200 transactions with a profit of 150 to be transferred. We don't know what it's attributable to because there are 1200 transactions. Unless you can attribute it, then it's a - it's misleading. Let's take 10 other transactions. Any incremental transaction becomes a three per cent sale, which is a nonsense. It can only be the case, if we can't allocate it - to me, if you can't allocate it it's actually a don't pass go point. We can't allocate the year-end TPA in the actual world and we cannot allocate the year-end TPA in the hypothetical world. How can you determine how much is attributable to the infringing products? There's no attribution. There is no attribution. It's necessarily a consequence of those 1200 transactions. You can't say 1000 transactions leads to 100 and 200 transactions leads to the 50 because that's not how it works. By definition we can't allocate it. 451 Mr Stone again responded: Sorry. By definition, if you've said that you start with the actual TPA being based on 1000 transactions giving rise to 100 and you're saying in the hypothetical world there would have been 1200 transactions and that would have given rise to a TPA of 150, then the difference between the actual and the hypothetical is 50 and it's caused by the difference in the actual to hypothetical change in number of transactions. By definition, it is calculated with reference to what has changed. I agree that it is not possible to sit there and say out of the original 1000 transactions how much of that 100 relates to each individual transaction because there are costs and overheads that relate to all of those transactions, but to the extent that those costs and overheads have been taken into account, the revenue has been taken into account from the extra 200 transactions, and when you crank the handle on the TPA without changing the mechanics of it, you end up with 50, then by definition, the 200 transactions have caused the 50. I don't understand why you then need to try and attribute those back to the 200 because you derived it from those 200. To my mind, I'm - I'm - as I said, I'm struggling to understand the distinction. 452 Moving to the topic of other discretionary expenditure in Australia unrelated to Lexapro and Esipram, Mr Stone agreed that he had assumed there was no such expenditure. He was unaware of any such expenditure. This evidence was given: MR HENNESSY: Just pausing there for a moment, is this broadly correct that in about - up until June 2000 the escitalopram products had accounted for something like 80 per cent of all products sold by Lundbeck Australia or CNS? MR STONE: I can't remember the exact percentage but it was a very significant percentage of that order of magnitude, yes. 453 Mr Stone was then taken to a letter dated 6 July 2009 from Lundbeck AU's Managing Director to all employees which said: As you are aware the escitalopram patent decision has changed the landscape dramatically for Lundbeck Australia. … …as we are coming from a lower sales base, the investments we can make in marketing our products must necessarily be reduced. As a consequence of the revised outlook for Lundbeck Australia we will be forced to make organisational changes in order to remain profitable and operationally sound. The details of these changes will be announced on the 14th of July at which time you will individually be informed whether your employment position will be impacted and whether Lundbeck will be able to offer you a position in the restructured organisation. 454 Mr Stone gave evidence in this exchange: MR HENNESSY: Doesn't that suggest to you that if the converse were to occur, that is, in a hypothetical world, if sales were to increase, that expenditure on marketing of products would also increase? MR STONE: Yes. MR HENNESSY: And that could be across the board, in other words, not just in relation to escitalopram products but other products being sold by Lundbeck Australia and CNS. MR STONE: The memo doesn't clarify which products it's referring to, so it may have included those; it may not have included those. The other products, that is. MR HENNESSY: Well, subject to that qualification, do you otherwise agree with my question? MR STONE: Yes. MR HENNESSY: But that's not a matter - that is, the possibility of expenditure on other products is not a matter that you took into account in determining the hypothetical year-end adjustment; correct? MR STONE: Correct. … I didn't see this as being a demonstration that there would have been expenditure on other products. … Reading it now, it would suggest that there is a possibility. Depends on how one reads that sentence. 455 Mr Samuel was also asked about this letter. He agreed it was concerned with letting employees go but did not agree that the letter was a thin basis for inferring that if sales of the escitalopram were greater than they were more would be spent on marketing other products, and that this concept would exist irrespective of the letter. He did not accept that the "investment" being referred to was investment in staff. He agreed that Mr Stone's calculations added back in the cost of additional staff in the hypothetical analysis of lost sales. He agreed that the issue was that he considered that if there had been increased sales of the escitalopram products then it was possible that Lundbeck AU might have had some discretion to spend some of the money in Australia rather than remitting it to Lundbeck A/S under the annual transfer price adjustment. 456 Mr Stone was asked about the exercise Mr Samuel had done showing that if the increased TPA was notionally allocated to Lexapro and Esipram sales then the gross margin would be different from those historically achieved on the sale of those products. Mr Stone said: The actual TPA was never allocated between products to my knowledge, and, therefore, an attempt to allocate the hypothetical TPA would seem to be irrelevant. It's not an exercise that was undertaken. …the irrelevance is that it's not done for the original, so doing it for the hypothetical doesn't make any sense. 457 Mr Samuel said: I agree it doesn't make any sense and yet that is actually what Mr Stone has done. He has allocated all of the adjustment to these infringing sales. He has done an allocation, having agreed that one shouldn't be done, and that's what this table is showing, is what the consequence of that would be. He has allocated the incremental margin in its entirety to the infringing sales and I'm saying you can't do that. A hypothetical year-end TPA is thousands of transactions. We don't know how it should be allocated.

19.3 Discussion 458 Sandoz made a number of arguments against Lundbeck A/S's claim based on the different year-end adjustments that would have been made but for Sandoz's infringing conduct (which would have resulted in increased payments by Lundbeck AU and CNS Pharma to Lundbeck A/S). 459 Sandoz's first argument was that if Lundbeck AU was (and is) the exclusive licensee of the 144 patent then "none of the losses that Lundbeck DN is alleged to have suffered are losses made in its capacity as patentee, and accordingly the quantum of Lundbeck DN's damage suffered as patentee is zero". Sandoz said that: The Applicants claim the following types of loss have been suffered by Lundbeck DN as patentee. First, the losses on sales to the Australian entities of escitalopram products that it has manufactured. Mr Stone identified the elements of Lundbeck DN's losses as being loss of gross margin earned on lost volume of Esipram sales, and lower transfer pricing year-end adjustment. It is apparent that the character of these alleged losses are those made by Lundbeck DN in its capacity as a foreign manufacturer of goods exported to Australia. Second, the Applicants now also allege that Lundbeck DN's losses for the entire period from June 2009 to December 2012, include the diminishment of the value of its shareholding in its Australian subsidiaries… there is no evidence proving or quantifying any loss to Lundbeck DN on the basis of a diminution of value in shareholding. However, on Lundbeck DN's own case, it has excluded itself from remuneration as patentee in Australia. It has done so by conferring a royalty-free licence on Lundbeck AU, which the Applicants claim in this proceeding is exclusive. An exclusive licence requires that the legal effect of the licence must be that the patentee and all other persons are excluded from exercising the right to exploit the patent. 460 Leaving aside the issue of the diminution in value of Lundbeck AU and CNS Pharma (which, as I will explain, it will not be necessary to address), I consider Sandoz's submissions misconceived. It may be accepted that because it was the exclusive licensee Lundbeck AU alone had the right to exploit and authorise the exploitation of the 144 patent in Australia. This does not mean, however, that Lundbeck A/S was no longer the patentee or that it was not entitled to claim for its own loss in infringement proceedings commenced by it in its capacity as the patentee. Sandoz cited no authority for this proposition. 461 In Black & Decker Inc v GMCA Pty Ltd (No 5) [2008] FCA 1738; (2008) 79 IPR 450 Heerey J awarded an account of profits to the patentee and exclusive licensee. In Roche Therapeutics, Inc v GenRx Pty Ltd [2007] FCA 83; (2007) 71 IPR 546, Emmett J considered whether the balance of convenience favoured the grant of an interlocutory injunction in favour of the applicants, two joint proprietors of a pharmaceutical patent and one exclusive licensee, referred to jointly as the "Patentees". The Patentees' pharmaceutical product was listed on the PBS. His Honour accepted that if the respondent and other competitors supplied the same pharmaceutical under the PBS the Patentees would suffer irreparable damage due to reductions in the price of their product under the PBS. Although his Honour did not distinguish between the proprietors of the patent and the exclusive licensee, these price reductions would have affected the proprietors of the patent irrespective of the identity from time to time of any exclusive licensee. He also said that if the respondent were found to infringe on a final hearing, "the Patentees would be entitled to damages for lost sales": at [103]. He noted that "the Patentees have also undertaken very substantial investment in research and development" which could only be recovered through sales of its products: at [102]-[103]. It appears this research and development was undertaken by the proprietors, whereas the exclusive licensee expended resources on educational and promotional activities: at [102]. It was literally a group effort and his Honour treated it as such. 462 In any event, the flaw in the proposition is exposed in Sandoz's statement, contrary to the fact, that Lundbeck A/S had "excluded itself from remuneration as patentee in Australia" by granting Lundbeck AU the exclusive licence to exploit the invention on a royalty free basis. This statement implicitly assumes that if Lundbeck A/S had granted the exclusive licence on the basis of payment of a royalty then Lundbeck A/S would be entitled as patentee to recover damages for the loss of the royalty payments. 463 Sandoz has not explained why any different principle would apply to the increased payments that Lundbeck AU and CNS Pharma would have made to Lundbeck A/S but for Sandoz's infringements. Considered in the context of the arrangements which were in place as between the members of the group of companies, it is simply inaccurate to say that Lundbeck A/S had excluded itself from remuneration as patentee in Australia. It had excluded itself from being able to exploit the invention in Australia, but that is a different matter from excluding itself from remuneration. It obtained remuneration from Lundbeck AU and CNS Pharma in the form of gross margins on sales and the annual adjustments to ensure that those companies retained a return on sales of approximately 3%. 464 There is no reason in principle which I can discern which would exclude Lundbeck A/S from recovering what it would have been paid by Lundbeck AU and CNS Pharma but for Sandoz's infringements. 465 Sandoz's next argument against Lundbeck A/S's claim was based on Mr Samuel's evidence. As noted, to the extent that Mr Samuel was considering whether the claim could be sustained at all, I consider the issue to involve a question of legal rather than accounting principle. 466 As discussed, the rights which Lundbeck A/S yielded to Lundbeck AU were all rights in Australia to exploit the invention claimed in the 144 patent including the right to authorise others to exploit the invention. The yielding up of these rights does not mean that Lundbeck A/S was no longer the patentee. It may be accepted that the arrangements between Lundbeck A/S and Lundbeck AU did not include payment of a royalty either in respect of Lundbeck AU's own exploitation of the invention or in respect of it authorising another entity to do so (which it did in the case of CNS Pharma). The reality is obvious. Lundbeck A/S did not need to require payment of a royalty because all three corporations were bound by the transfer pricing policy arrangements between them which had to be (and were in fact) consistently applied so the subsidiaries retained a rate of return on all sales of approximately 3%. This policy was continuously and consistently implemented throughout and on a year-by-year basis. 467 To my mind, the primary conceptual flaw in Mr Samuel's approach which has led him to the view that Lundbeck A/S cannot sustain its claim at all based on the additional sales that would have been made by Lundbeck AU and CNS Pharma but for Sandoz's infringements, is one of perspective. He is considering each lost sale of Lundbeck AU as an infringement of Lundbeck AU's rights and each lost sale of CNS Pharma as an infringement of CNS Pharma's rights. I do not consider this to be the correct perspective for current purposes. The claim is that of Lundbeck A/S as patentee. As patentee it suffered loss as a result of the fact that it did not receive the amount of money it would have received from its subsidiaries in the relevant years affected by Sandoz's infringements because Sandoz made sales that otherwise would have been made by Lundbeck A/S's subsidiaries. 468 As a matter of principle, provided a patentee or exclusive licensee proves loss caused by the infringements of a kind which is reasonably foreseeable, the object of an award of damages is to restore each of the patentee and/or the exclusive licensee to the position each would have been in had the infringements not occurred. I do not doubt that loss of the kind involved here, the remittal of a lesser amount of money from the subsidiaries to the parent as a result of lost sales caused by Sandoz's infringements, was reasonably foreseeable. I also do not doubt that the required causal connection is satisfied. The accountants agree that if (as I accept to be the fact) Sandoz made sales that otherwise would have been made by CNS Pharma and those sales (as well as sales of Lexapro) would otherwise have been made at higher prices (a fact which I also accept given Ms Weber's unchallenged evidence) then Lundbeck A/S would have received greater payments each year than it did receive under the transfer pricing arrangements of the Lundbeck group. 469 In my view, the circumstances are analogous to an arrangement for the payment of an annual royalty for the right to be exclusive licensee, despite not taking that form. The analogy, however, demonstrates why Lundbeck A/S's claim exists independently of any claim by Lundbeck AU and independently of the fact that CNS Pharma can make no claim itself for patent infringement. It is Lundbeck A/S's claim as patentee with which we are concerned. Further, there can be no double claim because to the extent that Lundbeck AU's claim is subsumed by Lundbeck A/S's claim, it is the latter which should prevail because that will reflect the position that would have existed but for Sandoz's infringements. Given that the object must be restoration, insofar as possible by an award of damages, that object is best achieved by recognising what the position would have been if not for the infringements - which is increased money being received by Lundbeck A/S. 470 In common with Mr Stone, I have no concern about the issue of allocation to different products. This is not how the arrangements worked in practice. There was never any such allocation. As a result, Mr Samuel's exercise in attempting to prove that gross margins on Lexapro and Esipram would be distorted is immaterial. The material fact is that there is no dispute between the accountants about the amounts that Lundbeck A/S would have received assuming the lost sales and the decreased prices. It is those amounts which represent Lundbeck A/S's loss for which, as a matter of principle and fact, it should be compensated by Sandoz. 471 For these reasons I do not accept Sandoz's submission that to the extent its claim relied on the year-end transfer pricing adjustment it would have received, Lundbeck A/S's claim is unsustainable. It does not matter that the adjustment relates to all products and not just Lexapro and Esipram. What matters is that it is clear from Mr Stone's approach that the additional amounts Lundbeck A/S would have been paid can only relate to Lexapro and Esipram because those amounts reflect only the lost sales and/or decreased prices of those products and nothing else. It is not the case that "neither Mr Stone nor Mr Samuel were able to properly calculate the counterfactual end of year financial adjustment, because there was no evidence led by the Applicants as to the counterfactual transfer prices of Lexapro, Esipram and all other products sold by Lundbeck AU and CNS Pharma, and no evidence of the counterfactual expenses of Lundbeck AU and CNS Pharma during the counterfactual period". This submission fails to confront the actual circumstances which would have applied if not for Sandoz's infringements. Lundbeck A/S has proved what it needed to prove as patentee - that if not for Sandoz's infringements it would have received more money than it did from its Australian subsidiaries (the exclusive licensee exploiting the invention and the subsidiary of the exclusive licensee authorised by it also to exploit the invention). 472 As I have said, conceptually, I see no material difference between the circumstances of Lundbeck A/S not receiving the amount of money it would have received each year if not for Sandoz's infringements and a patentee not receiving an annual royalty payable by its exclusive licensee for its own exploitation of a patent and its authorisation of a patent by any other person. It follows that I do not accept the materiality of Sandoz's submission that "by reason of the terms of the Applicants' transfer policy and as agreed by both experts, no particular lost sale would have resulted in a transfer adjustment". It is not being suggested that there would have been merely a single lost sale. Further, it is necessary to consider the kind of loss Lundbeck A/S is claiming. The loss is the loss of the money it would have been paid annually by its subsidiaries pursuant to a consistently applied and documented transfer pricing policy if not for Sandoz's infringements. As a matter of principle I do not see such loss as novel or as existing at the penumbra of the kinds of reasonably foreseeable loss a patentee such as Lundbeck A/S may well suffer and, on the facts of the present case, has proved that it did suffer. 473 I also do not see the circumstances of this case as giving rise to the kind of issues considered in Gerber Garment Technology Inc v Lectra Systems Ltd [1995] RPC 383 and [1997] RPC 443. In the present case, the relevant loss is that of Lundbeck A/S itself as patentee. To the extent that the subsidiaries of Lundbeck A/S had to remit money to Lundbeck A/S to maintain an annual rate of return of sales of approximately 3%, the loss is not that of the subsidiaries but of Lundbeck A/S (which is how Mr Stone has accounted for the claims). In Gerber it was the subsidiary which suffered loss. Mr Justice Jacob characterised the patentee's loss as a loss in value of its shares in its subsidiaries (at 409). The UK Court of Appeal held that while such damages could be recovered as a matter of principle, the patentee in Gerber had not proved that it suffered such loss. Hobhouse LJ at 479 said: The root principle which must be adhered to is that each company is a separate legal entity. The property of one is not the property of another. The plaintiff must prove its own financial loss in its own pocket and quantify it. Any other approach is contrary to the decided authorities and the principle in Saloman v A Saloman & Co Ltd []. The plaintiffs have not attempted to discharge this burden of proof. 474 Lord Justice Hutchison at 482 said: I have no difficulty in accepting that, subject to proof of damage, the patentees can recover the losses they have suffered by reason of the diminution in value of their share holding in or dividends from the subsidiary companies brought about by the infringements. The vital question remains as to whether the judge was entitled to hold that they had proved their loss. There can be no doubt that the onus which rests on a plaintiff in relation to proof of damages requires, in such cases as the present, that the plaintiff company should establish that it has suffered damage by reason of the losses suffered by its subsidiary company. It is not enough simply to demonstrate that the profits of the subsidiary have been diminished: the plaintiff company can only recover in respect of its own loss. The critical question, it seems to me, is how such an onus can be discharged. 475 Lundbeck also ran a Gerber style argument (loss in the value of its subsidiaries) as an alternative to its principal argument based on loss of the payments it would have received as part of the transfer pricing arrangements between members of the group. I have accepted Lundbeck A/S's principal arguments. It has proved its own loss. If my conclusions in this regard are incorrect, I would consider that Lundbeck would have to contend with the same problem that confounded the claim in Gerber. However, Lundbeck has confronted this problem through the accounting evidence. Mr Stone said: So CNS is a wholly owned subsidiary of Lundbeck Australia. So to the extent that CNS is out of pocket, then, effectively, the value of that investment by Lundbeck Australia... is equally diminished by the extent to which they are unable to claim that loss and, therefore, there's a diminution in value. At that point in time, or certainly in 2010, there were dividends flowing from CNS to Lundbeck Australia. So I would expect that, to the extent that they had had more moneys available to them, even after the TPA, that those funds would have likely flowed to Lundbeck Australia along with the dividends that did in actual fact flow through that period. 476 Mr Samuel responded that he did not disagree "about the value question" but said it was not clear to him if it was a loss of dividends or of value which might lead to different tax treatments and different outcomes. This evidence was given: MR BANNON: But taking it step by step, am I right in thinking you would accept that it's a loss of value of Lundbeck Australia if CNS can't recover the money for whatever reason. MR SAMUEL: There would be a loss of value, yes. MR BANNON: Equal to the - equal to the money that CNS can't recover. MR SAMUEL: No, equal to the after-tax consequences of the money it can't recover. MR BANNON: In CNS's hands or in MR SAMUEL: CNS would have to - well, actually, this is another question I - on reflection, I think they are a consolidated tax group, so again there's another tax question. MR BANNON: Yes. MR SAMUEL: Conceptually, they would have to pay tax on any income they would make. You can only pass up by way of dividend and after-tax amount. You can't declare a dividend out of a pre-tax amount. So the loss of value would be equivalent to the after-tax figure and then there's a question of whether you would benefit from franking credits in the holding company, and we're into a tax world I'm not entirely familiar with. MR BANNON: But I think you do know that if they do provide tax returns on consolidated basis or MR SAMUEL: I think. That's my recollection. MR BANNON: Yes. Sorry. Does that affect your answer? MR SAMUEL: Sorry. Sorry. Let me answer that - retract that. I understand they operate as a consolidated tax group. I don't know how they provide tax returns. MR BANNON: But does that affect the answer you gave or were you answering on that basis? MR SAMUEL: I'm answering on the basis that conceptually they would have to pay tax on the profit, so it's not clear if you're looking at the value of an investment in the subsidiary - sorry. Just - my expectation would be that it's the after-tax value that you've lost, not a pre-tax value. MR BANNON: But if it's - if they operate as a consolidated unit and provide returns on a consolidated unit, then it would be a question of tax in Lundbeck Australia's hands, wouldn't it? MR SAMUEL: I don't know the answer to that. MR BANNON: Mr Stone, have you got any - can you add to that discussion? MR STONE: My understanding would be that they would - if they were to pay dividends, which I would agree would be an after-tax - out of after-tax profits, but it would have a franking credit that went with it. So effectively when it receives in the hands of Lundbeck Australia, Lundbeck Australia would receive the after-tax amount but would then have to pay no further tax because the franking credit would exist. So in my example where $2 million was out of pocket in - in CNS, they would then pay tax on that amount. If they paid that amount in full as a dividend, then the amount received by Lundbeck Australia would be the after-tax amount, and then it's a question - and I agree with Mr Samuel - it's a question of whether or not that would be treated as effectively income or capital. If it's treated as income, then it would be taxable in the hands of Lundbeck Australia, and, therefore, to get the 1.4 million in my example, they would have to have the full $2 million. If, in fact, it's treated as some sort of capital amount, then that would become a more complicated tax question and I'm not sure what the answer is to that. MR BANNON: Is it the case that at worst the after-tax value of the extra receipt would be a loss to Lundbeck Australia? MR STONE: Correct. MR BANNON: But it may be a grossed-up amount. MR STONE: Correct. MR BANNON: Mr Samuel, any different view? MR SAMUEL: I was just going to add if it's a dividend, there might be timing issues on when that would be paid, so the loss might not crystallise in Lundbeck Australia until such time as a dividend would, in fact, have been paid. MR BANNON: Would you agree at worst the loss of value of Lundbeck Australia would be the after-tax value of the return? MR SAMUEL: Yes. 477 If it were necessary to so conclude I would be satisfied that this evidence is sufficient to establish the dollar for dollar loss in value of Lundbeck AU's shares in CNS Pharma. 478 To the extent that Mr Samuel was concerned that if Lundbeck AU and CNS Pharma had made increased profits of the magnitude involved in the agreed calculations, then they might have spent more money on other products, I do not consider that the letter of 6 July 2009 provides a reliable evidentiary foundation for inferring that this was likely. The letter, read as a whole, is concerned with the redundancies that were going to have to be made given the loss of exclusivity over escitalopram products in Australia. 479 Contrary to Mr Samuel's view, I am not persuaded that it stands to reason that if Lundbeck AU and CNS Pharma would have made substantially more sales of their major products (Lexapro and Esipram) they would spend more money marketing other products. 480 I also accept the following submissions for Lundbeck: (1) Mr Andersen rejected any possibility of such expenditure instead of remittal to Lundbeck A/S as required by the transfer pricing arrangements between the companies. He described the exercise as a "fairly mathematical effort" which was adhered to by adjusting at year-end as required. (2) Mr Andersen indicated that if special expenditures were required those expenditures would be funded by Lundbeck A/S rather than from local funds because of the need for consistency and transparency in the transfer pricing arrangements. (3) Sandoz did not cross-examine Darren Murray who was Lundbeck Australia's Commercial Manager. As Lundbeck put it: It was squarely within Mr Murray's role to be aware of the types of expenditure that Lundbeck AU and CNS Pharma incurred and considered during the relevant period. Sandoz did not put to Mr Murray that speculative expenditure of the type contemplated by Mr Samuel might have been incurred by Lundbeck AU during the relevant period. Sandoz' failure to cross-examine Mr Murray about this issue deprives Mr Samuel's speculation of any force. (4) Mr Samuel's argument about the distortion of gross margins on Lexapro and Esipram does not involve a faithful application of the inter-company arrangements because "the transfer pricing policy is not applied on a product-by-product basis. It follows that his analysis of the gross profit margins of individual products is inapposite and distracts focus from achieving an overall ROS of 3%. Further, the transfer pricing policy is not concerned with gross margin, but rather net margin". 481 Overall, I would assess the possibility of expenditure on other products in Australia when the money should have been remitted to Lundbeck A/S as approaching the end of the scale of possibilities which is merely speculative. I will weigh it in the balance when considering any further adjustment for risk having regard to all relevant circumstances but the weight which should be attributed to it, in my view, is very slight. 482 As such, I do not accept Sandoz's submission that Lundbeck A/S's claim based on the money it would have received each year from its subsidiaries is defeated because Mr Stone's evidence is based on the unfounded assumption that those subsidiaries would not have retained some profits and spent them on other products rather than remitting to Lundbeck A/S as required by the transfer pricing arrangements. As I have said, this is a bare possibility, approaching the speculative, because on the evidence: (1) the companies consistently acted in accordance with the transfer pricing arrangements, acting inconsistently with the transfer pricing arrangements would have jeopardised the tax position of the group; (2) there is no evidence of such an event; and (3) it is inherently unlikely that increased sales of Lexapro and Esipram would have resulted in increased expenditure on unrelated products (given that the vast majority of the businesses involved Lexapro and Esipram sales). 483 When all of these matters are considered there is no justification for giving this issue material weight in the overall analysis. I do not disregard it, but the weight to which it is rationally entitled is very slight.

20. CNS Pharma claim period 484 It will be apparent from the above that I consider that the dispute about the CNS Pharma claim period is a non-issue. CNS Pharma did not commence proceedings until 30 May 2016. CNS Pharma's claim concerns alleged misleading and deceptive conduct. Sandoz contended that CNS Pharma could not maintain any claim before 10 December 2010 because of the statutory limitation period. Because CNS Pharma cannot make a claim for patent infringement, the issue does not arise in respect of Lundbeck A/S's patent claim. However, to avoid any doubt, it is appropriate to record my view that insofar as Lundbeck A/S's patent claim is concerned, there is no issue with respect to CNS Pharma. As discussed above, the claim is that of Lundbeck A/S as patentee for loss it has suffered as a result of Sandoz's infringements. That loss includes the loss of the money that CNS Pharma (as an authorised exploiter of the 144 patent in Australia) would have paid Lundbeck A/S as the patentee irrespective of the fact that CNS Pharma cannot bring a claim for patent infringement. No limitation period with respect to CNS Pharma can affect the claim of Lundbeck A/S. 485 Mr Stone proceeded consistently with the conclusions above. He dealt with the transfer pricing adjustment on the global basis which would have occurred but for Sandoz's infringements. He did not confine Lundbeck A/S's loss by reference to any limitation period affecting CNS Pharma by reason of the fact that it did not commence proceedings until 30 May 2016. This approach properly reflects the principle that the object of an award of damages for patent infringement is to place the patentee (and exclusive licensee) in the same position as each would have been in if not for the infringements to the extent a monetary sum may do so.

21.1 The issue 486 As Lundbeck conveniently summarised: Mr Stone and Mr Samuel disagree as to how certain overheads should be treated in the assessment of loss. The overheads in question include both direct costs tracked specifically to LEXAPRO and ESIPRAM and indirect costs which are not tracked to those products. Mr Stone has identified incremental overheads of $44.751 million which are to be deducted from the loss suffered by Lundbeck. In the joint evidence session, Mr Stone amended that assessment to agree with Mr Samuel's assessment of additional distribution costs which adds an approximate $4 million to the incremental overheads bringing the figure to $48.751 million. In contrast, Mr Samuel has identified incremental overheads of between $62.924 and $66.557 million (depending on which counterfactual scenario is adopted) which are to be deducted from the loss suffered by Lundbeck. Of course, these figures are not specific to Sandoz, but rather concern incremental overheads in respect of all lost sales.

21.2 The accounting evidence 487 As explained in the accountants' joint report: Mr Stone had reference to a budget document for the calendar years 2009 and 2010, which, he was instructed, was prepared prior to the Four Generics' entry into the escitalopram market (the Budget). Because the Budget was prepared prior to the Four Generics' entry into the escitalopram market, Mr Stone considered this a reasonable estimate of counterfactual overhead costs. 488 As further explained in the joint report: In Mr Samuel's opinion, there are three methodologies which can be applied, depending on the circumstances, for estimating hypothetical overheads arising from a change in volume, being: a) A "ground up approach", which investigates closely the nature of each cost for the purpose of determining whether it would have varied with the assumed change in volume. This methodology has not been adopted by either Mr Stone or Mr Samuel in this proceeding; b) A "top down" approach which analyses the movement in overheads relative to the movement in volumes. This is the approach Mr Samuel has adopted in this proceeding; and c) Reliance on budgets, which analyses the movement in overheads based on contemporaneous budgets. This is the methodology adopted by Mr Stone in this proceeding. In summary, Mr Samuel's approach was to estimate the variable total overhead cost per incremental unit after removing distribution costs (which are accounted for separately) and certain fixed costs (such as rent and registration). Mr Samuel considered this approach to be appropriate as: a) having reviewed OS1 and enquired during inspection at Lundbeck as to the level of detail supporting the budget on which he relied, concluded that the budget was not a reliable basis for estimating the hypothetical overheads that would have been incurred; b) insufficient evidence was provided to enable a ground up approach, and Mr Stone had not attempted such a methodology; c) the incremental increase in hypothetical volumes was very significant. The escitalopram products represented 79% of volume prior to the infringements/contraventions and up to 89% of the hypothetical unit sales during the loss period. It follows that the CNS/Lundbeck Australia business, including all fixed overheads, are fundamentally a function of the escitalopram products. A significant change in volume, as is claimed in all the scenarios, can reasonably be expected to impact on all costs that a have a variable element; and d) CNS did not record any overheads. Accordingly, it was necessary to analyse overheads for both companies on a consolidated basis. 489 In response, Mr Stone said this: Whilst both approaches necessarily involve a degree of estimation, Mr Stone considers his approach, which involves a specific consideration, based on contemporaneous estimates, of particular cost types is better than Mr Samuel's more mechanistic approach, which does not consider the potential impact of the hypothetical scenarios on particular types of expenditure. 490 In their concurrent evidence session the accountants explained their approaches. 491 Mr Stone confirmed that he had accepted Mr Samuel's figures in respect of distribution costs, so that is now a non-issue. 492 Mr Stone said that he understood the budget had been prepared by Mr Murray (which is a fact) and that Mr Murray is a chartered practising accountant (which is also a fact) and assumed that Lundbeck is a "significant and sophisticated company" (an assumption which is supported by the evidence). He understood (correctly) the budget to be showing likely sales and likely costs assuming no infringement of the 144 patent in 2009. He accepted it was "a real time estimate by people who you would have assumed were familiar with the actual costs" who were "equally familiar with any variances they had in the past from prior budgets and actual costs". 493 Mr Samuel, for his part, accepted that the budgeted costs might be lower than provided for in the budget. He said he had discarded the budget because: ….It was a variance of up to 39 per cent, as I recall, in certain cost categories, and also because no detail - no intelligible detail was provided that would enable me to assess how that budget was put together and, therefore, whether it was based on reasonable assumptions or not. 494 Mr Samuel explained that as a result of his rejection of the budget: …my approach was to look at overheads as a whole. … my approach starts from the premise that we have a structural change in the business that relates to two products that represented nearly 80 per cent of volume prior to the infringements and as much as 89 per cent of the volume in the hypothetical scenarios to the extent that we had a 230 per cent variance between actual and hypothetical in one of those years. So we had a - we have a whole of business consequence as a result of the infringements, and, therefore, my approach is to say that, well, in effect, the heads are going to be impacted by - by that sort of massive structural exchange. … …it's a whole of portfolio approach. It's not a ground-up cost by cost approach. Simply - that simply isn't what I've done. So it looks at overheads as a whole and sees how they performed as a - as a portfolio or as a bucket when volumes have moved from one year to another. 495 Mr Samuel accepted that in his whole of portfolio approach he had treated tax consulting and legal fees as a variable cost (a treatment with which Mr Stone disagreed). Mr Samuel defended his approach, saying: … There will be some overs and unders in my approach at individual cost lines. Some costs will be included incorrectly, some costs will be excluded incorrectly in the sense of how they're moving. It's a whole of portfolio approach, and, frankly, I think it's the only way you can do it when you don't have the detail to do a ground-up approach and you don't have a budget that's reliable. 496 Mr Samuel agreed he had used the most recent year before the infringements (2007-2008) as the basis for his incremental increases in costs, which he considered reasonable because of the increase in the number of units sold in that year. 497 Mr Samuel agreed he had used the increase in promotional costs between 2007 and 2008 as the basis for an increase in such costs in the hypothetical analysis and thus also treated that as a variable cost. Mr Samuel said because of his approach he was not seeking to justify any cost as variable or not. The same applied to subscription costs, and advertising costs. 498 Mr Stone in response said this: I appreciate that budgets are necessarily an estimate of what would have incurred but the estimation process is writ large within the estimation of damages, so any of the approaches that Mr Samuel suggests or I've used are necessarily estimates. From my perspective, it is more likely that somebody sitting there at the time doing a budget about the hypothetical world is going to come up with an estimate that is a reasonable estimate. I've got no reason to suggest that it's unreasonable. The fact that there are variances on individual lines within a budget is not unusual. The - the fact that the - overall the budget ends up with costs for the main relevant areas that are within one to eight per cent of reality suggests that there was a reasonable process. … But it is just that; it's an - it's an estimation. But in regard to the budgets, they seem to be a reasonable starting point. They're contemporaneous. They were done with the hypothetical world in mind at the time by people who understood the business. 499 In response to some specific items, Mr Stone gave this example: …I would come back to this subsidiary expenses defrayed by headquarters. The idea that you have an increase in a cost of one and a half million dollars because that's the effect of taking out the credit, is to increase costs by one and a half million dollars, when you can see that at least 700,000 of that relates to a cost that is one-off seems to me to be flying in the face of the idea that these costs are likely to be variable. I understand Mr Samuel's point that says he has taken up a sort of holus-bolus approach to this, but when you look at that, we do have sufficient detail in the trial balance to be able to raise questions about items that are likely to be fixed. The effect of Mr Samuel's calculation is that all the of the variance in overheads, regardless of its reasons, is included as being variable. So if a cost had increased simply because they had done one more study that year, it was fixed, it was never going to be done again, Mr Samuel's figures necessarily take that cost and apply it per unit going forward. To my mind, that doesn't make sense when you can see that these costs appear to be one-off. Especially, I take Mr Samuel's point that his calculations, albeit with the way that he has treated the defrayment, get similar numbers for the '06/07 comparison to '07/08, but, conversely, the '05/06 comparison gets a negative variance on an overall basis. That, to my mind, highlights the problem with trying to take this top-down approach without regard to the detail in looking at the way in which costs are likely to move. Necessarily within the overheads are costs that are going to be fixed. There are some that will go up each year, there were some that go down. And I accept Mr Samuel's point that he has taken the downs as well as the ups, but he has taken the downs and the ups for costs that are necessarily must have been one-off or fixed in nature. I don't know exactly which each of the items that were fixed and one-off are, but I can see some clues from the trial balance, and that trial balance suggests to me that the top-down approach that he has done, albeit now adjusted for certain items - at the moment he seems to be sort of half top-down and half bottom-up, no regard for the budget, and doesn't actually fully take account of a major variance which seems to be the defrayment of costs. 500 Mr Samuel accepted that Lundbeck's budget showed promotional costs decreasing between 2009 and 2010 despite sales of escitalopram overall being projected to increase in that period. He accepted also that promotional costs involved a discretionary decision. This exchange then occurred: MR BANNON: And you understand that promotional costs with an innovator company fundamentally related to the number of people you employed to go and visit doctors. Do you understand that? MR SAMUEL: Yes, I expect so. MR BANNON: You understand that - do you understand that innovator companies like Lundbeck, they don't promote to pharmacists? You understand that? MR SAMUEL: Well, that's not always correct. In this instance, they ended up with a mixed distribution model in which they did go direct to pharmacy. MR BANNON: But that's not a promotion of a product. It's a distribution model, isn't it? MR SAMUEL: I seriously doubt you could distribute to the pharmacies without promoting to them first. MR BANNON: Well, do you understand this is a prescription product for depression? MR SAMUEL: Yes. MR BANNON: There's no promotion you can do to increase the incidence of depression in the community. MR SAMUEL: No, but you need to establish relationships with the pharmacies in order to distribute to them directly. MR BANNON: Yes, but the pharmacists - you understand that the pharmacists will only write the scripts given to them by patients who get them from doctors. MR SAMUEL: Yes. MR BANNON: And the only promotion anyone can do of a product like this, absent the generics out of the market, is promotions to the doctors. You understand that? MR SAMUEL: No. Well, I don't know if that's the case or not. 501 Mr Samuel accepted also that the number of people employed in the administrative department was a discretionary decision for Lundbeck. This exchange then occurred: MR BANNON: So insofar as you say budget item, for example, for total promotion costs and admin department and marketing department is an unreliable estimate, you're, effectively, saying, aren't you, that based on your opinion the court should regard statements by a company of how much they, effectively, intend to spend as, what, to be doubted that they - that they won't spend that. Is that right? I'm just trying to understand what is the expertise you're bringing to bear to invite the court to disregard, in effect, what's an intentional statement by the company. MR SAMUEL: My expertise, your Honour, is as an accountant and a valuer. I look at forecasts and budgets all the time, every year. And what we have here is circumstances in which the budgets have been demonstrably wrong. And we have - I made numerous requests for the information to support these assumptions and nothing useful was provided. So I have no basis independently to say that this is a reliable budget, absolutely none. At the high level we know that they've got these line items that you brought to my attention wrong by material amounts by up to 39 per cent. 502 Otherwise, Mr Stone accepted that there was a risk inherent in his estimation of overheads and what he described as a minor risk inherent in his assessment of the costs of goods sold, as well as a risk in relation to his reliance on the budget. As he put it, this is all true for all calculations of the kind required, including those of Mr Samuel. Mr Stone said he could not comment on the percentage discount that might be applied to account for these risks. Mr Samuel effectively agreed with these observations. Mr Stone said this: I think it would be fair to say that if I'm looking at Mr Samuel's approach, then I think the estimation is likely to be overcooked. If I'm looking at mine, it potentially has some undercooked element. But ultimately there's an estimation on both sides, so both of them could be up or both of them could be down. 503 Mr Samuel said: … Where there are hypothetical cash flows, there are risk - there is risk. The risk of overstatement or understatement doesn't obviate the need for discount. Valuers, when they put together cash flows, always put together their best estimate of the most likely outcome. That means it could go up or down but you discount them anyway. 504 Mr Stone responded: The only comment I would make is that valuers tend to be looking at forward-looking estimates and that they're trying to work out what somebody is prepared to pay to take on that risk. To my mind, estimating loss and damage when you're looking historically is fundamentally a different exercise than forward-looking valuation exercise. But I agree with Mr Samuel, if one is taking a forward-looking valuation, then yes, there is a discount for risk. Whether that is appropriate in these circumstances where you're estimating historical loss and damage is, as Mr Samuel, I think, alluded to, a legal question.

21.3 Mr Murray's evidence 505 Mr Murray was employed by Lundbeck AU between 2003 and 2016. He is a certified practising accountant. Between 2008 and mid-2011 he was Lundbeck AU's Commercial Manager and from then to 2016 he was its Commercial Director. He is now a consultant to Lundbeck AU. 506 In his affidavit Mr Murray explained that he had prepared the budget for 2009/2010. He did so assuming that Lexapro and Esipram would be the only escitalopram products on the market and that the growth trend of recent years in the anti-depressant market would continue. As Lundbeck said, he explained as follows: 41. The process I followed to prepare the sales forecast for this budget revision was as follows: (a) I started working from the most recent previous sales forecast I had prepared (in or around April 2009). (b) I updated the forecast so it included actual sales for the period January 2009 - April 2009; (c) I updated sales volume forecasts for May 2009 to December 2010, taking into account the recent sales volume trend for escitalopram (based on Lundbeck Australia's sales data), the recent anti-depressant market growth trend (based on IMS Health data), and the additional sales volume expected from the engagement of the external contract sales team (PharmaLink) on around 2 March 2009 (based on the business case supporting the decision to hire the team); (d) I reviewed selling price forecasts based on my knowledge of any impending price changes; (e) I calculated new sales value forecast for 2009 and 2010, and compared this forecast to earlier budget forecasts; (f) I presented the draft to Klaus Abel. 42. Based on the forecasts in the budget revision document, in May 2009 my expectation was that escitalopram sales would experience annual growth of about 7% in 2009 and about 8% in 2010. These growth forecasts were higher than earlier forecasts made in September/October 2008, because of stronger than expected growth in the antidepressant market during January-March 2009 (-15% growth, compared to budget assumption of 6%), and due to the engagement of the contract sales team (PharmaLink) from around 2 March 2009. 507 Mr Murray also gave evidence that after the launch of the generic escitalopram products in June 2009 sales of Esipram fell sharply. However, overall sales of escitalopram products continued the trend of consistent growth from 2005 to 2013. 508 Sandoz did not cross-examine Mr Murray. It did not put to Mr Murray any of the concerns Mr Samuel expressed about the reliability of the budget.

21.4 Discussion 509 I find Mr Stone's overall approach to the issue of overheads persuasive. 510 In particular, I do not consider Mr Samuel's rejection of the 2009/2010 budget as a starting point for the required analysis to be sound. Mr Stone recognised that the budget represented a contemporaneous document, produced by the entity best placed to know its own business, assuming that the generic escitalopram products did not enter the market in June 2009. This reflects the foundational requirement of the hypothetical analysis which must be carried out in the present case to determine what loss Lundbeck has sustained by reason of Sandoz's infringement of the 144 patent. Mr Stone's approach also recognised that as he and Mr Samuel were engaged in a process of estimation there was no sound reason to infer that his starting point for the process would be more reliable than that of Mr Murray who was the person within Lundbeck AU responsible for the preparation of the budget, a document on which Lundbeck AU had been willing to rely for its own commercial purposes. Indeed, to the contrary, Mr Stone appreciated the fact that the budget was a contemporaneous document created for a serious commercial purpose by a person who, given his qualifications, experience, role and responsibilities, meant that the budget would be the best evidence for the start of his hypothetical analysis. I consider that Mr Samuel's approach, to disregard the budget altogether because of variances between budgeted expenditure and actual expenditure in the most recent year, was unreasonable in the circumstances. 511 As Lundbeck submitted: It was open to Sandoz to challenge Mr Murray as to the process by which the budget was prepared and the robustness of the forecasts it contains. Sandoz elected not to cross-examine Mr Murray at all, let alone about these matters. Further, the variances which are evident in a comparison of the 2007 and 2008 budgets are modest, and suggest (as Mr Stone stated) that Lundbeck AU's budgeting is reliable. Mr Samuel agreed in cross examination that he assumed that the preparation of the budget took account of prior year variations. As submitted above, the contemporaneous budget assessments are a peculiarly valuable piece of evidence for the Court. Mr Samuel's attempt to discredit the estimates included what he regarded as unsatisfactory responses to requests for documents underlying the budget. But the nature of the requests must be examined. A current absence of working documents in 2017-2018 in relation to a budget estimate created in 2009 is hardly surprising. Estimates by their nature have their embodiment in a final document. There is no logical reason for working documents or drafts or calculations to be kept. The documentary requests make the failure to ask Mr Murray any questions about the budget estimates even more telling. Further, during his oral evidence, it became clear that Mr Samuel's reservations about the budget were not based on any knowledge of defects in the way it was prepared. Rather, his concerns are based on the possibility that the budget was not prepared in a rigorous and careful manner. Such general and uncertain concerns are insufficient to cause the Court not to accept a contemporaneous budget as reliable. Further, Mr Samuel acknowledged that the budget set out the expenditure that Lundbeck AU intended to incur (exercising its knowledge of the market and its commercial judgment as to the appropriate level of expenditure) during the relevant period. 512 I have other concerns about Mr Samuel's approach to overheads which lead me to the view that his calculations are far more likely to over-estimate expenses than Mr Stone's calculations are to under-estimate expenses. 513 First, I do not accept his "structural change to the business" position. Escitalopram products were always the vast majority of the business. There had been steady growth in that business since 2005. Lundbeck AU (and its subsidiary CNS Pharma) had managed this annual growth organically without any "structural change". Until the decision of the Full Court against Lundbeck the companies anticipated continued organic growth. The fact that the entry of generic escitalopram products into the market in June 2009 had such a large impact on sales of Esipram in particular (so that, but for generic entry, there would have been many more sales) does not mean that in the hypothetical analysis it should be assumed that the business would have experienced a structural change. Accordingly, the justification for the "top down" approach which Mr Samuel used does not exist. 514 Second, I do not accept that it was reasonable to base the kind of increases in overheads which Mr Samuel's calculations yielded on the variances between 2007 and 2008. As Mr Stone noted, there had been a decrease between budgeted expenses and actual expenses between 2005 and 2006 which Mr Samuel disregarded. Rather, Mr Samuel had "taken one of the highest Variable Overhead Costs per incremental unit over the three years for which he compares the increase in actual units to the 'increase in actual overheads'". Further, the sheer magnitude of the increases in overheads posited by Mr Samuel indicates that a more cautious approach was required than his "top down" approach. 515 Third, the fact that there are so many costs in Mr Samuel's approach which have been inflated but which, on any reasonable view, would not be variable costs affected by an increase in sales also indicates that his calculations are likely to be unreliable. His defence of his approach, that some costs will be an over-estimate and some an under-estimate, so "cherry picking" is a pointless exercise, was not persuasive. In particular, his approach of including the large one-off cost that was defrayed by Lundbeck A/S for a study, in my view, appears unreasonable. 516 Fourth, Mr Samuel's approach to promotional costs did not appear to appreciate the fact that the sales of escitalopram had been consistently increasing over years with a decrease in promotional costs, or that the only promotion possible of the sale of the products themselves is by persuading doctors that the product should be prescribed to patients. As Lundbeck submitted: Furthermore, there was ample evidence that promotional expenditure by Lundbeck was not a variable cost required to produce the hypothetical escitalopram sales which would have been achieved in the Initial and Springboard Periods. The hypothetical sales are the sales in fact made by the generics generally and for present relevant purposes, Sandoz in particular. The overall market for escitalopram continued to increase after 2009 notwithstanding the fact that in 2009, after the generic launches, Lundbeck AU reduced its promotional activities and laid off promotional and marketing staff. Yet despite that, sales of escitalopram overall continued to increase. That also reflects the evidence of Mr Turner to the effect that Sandoz estimated in June 2009 that the total escitalopram market was growing in size in terms of units sold which evidence was not stated to be dependent on promotional activity by Lundbeck. Similarly, Mr Murray's unchallenged evidence is that the growth in the escitalopram market was stable before June 2009 and, with the benefit of hindsight, Mr Murray has identified that the forecasted trend was accurate. The projected costs, including promotional expenditure, were set with regard to that forecast. Both Mr Murray's and Mr Turner's projections followed market trends. Mr Samuel's oral evidence was that he did not consider the forecasted sales in the budget. Hence, Mr Samuel's untutored assumption that increased promotion by Lundbeck would have been required to generate the hypothetical sales of escitalopram is contradicted by not only the budget estimates, but also the objective evidence and Sandoz own assessment. It is to be noted that the generics do not promote escitalopram to doctors. In other words…the infringing sales were not generated by promotional work of the generics in persuading doctors to write prescriptions. The sales were a function of medical perception of the utility of escitalopram in the treatment of depression, no doubt enhanced by prior promotional work of Lundbeck. The cross examination of Mr Stone to the effect that keeping promotional costs at a fixed level was unrealistic proceeded in apparent oblivion to the objective evidence referred to above and Sandoz' own corporate evidence on the subject. 517 I also accept Lundbeck's submission to this effect about promotional costs: Moreover, the Budget demonstrates that Lundbeck was not planning any promotional expenditure on Esipram. That is reflected in the fact that the fuller form of the budget which shows prior year expenditure demonstrates that Lundbeck did not spend any money promoting Esipram. This is understandable because Esipram was sold by Lundbeck at a lower price and it made more money from Lexapro sales. Despite the lack of promotion of Esipram, were continuing to increase as observed by Mr Turner and had the majority market share in 2009. Lundbeck does not claim any lost sales of Lexapro. The only lost sales were those of Esipram. The objective evidence is that not a single dollar of additional promotional expenditure was required to generate the hypothetical sales of Esipram. None would have been spent by Lundbeck to generate the sales and none was required to generate the sales. Hence, the inclusion of any amount of incremental expenditure for promotion is not required and its inclusion understates considerably Lundbeck's claim. 518 Fifth, Mr Samuel appeared to give no weight to the fact that Lundbeck AU was in ultimate control of the kind of overheads which Mr Samuel was considering. The fact that there had been variances in the past does not mean that Lundbeck AU was at the mercy of external forces for these kinds of expenses. 519 For these reasons, I consider Mr Samuel's approach to overheads to involve a high likelihood of a substantial over-estimate. I do not accept Sandoz's contrary submissions. Those submissions do not take account, or sufficient account, of: (1) The forensic effect of Sandoz deciding not to cross-examine Mr Murray about the budget. (2) The fact that the budget is a contemporaneous document prepared for a serious commercial purpose by a person with the requisite qualifications, experience and responsibility to ensure that the document was fit for purpose. (3) The unreasonableness of Mr Samuel's outright rejection of the budget based on variances between the two most recent years which included one-off costs. (4) The high risk of over-estimating the overheads inherent in Mr Samuel's "top down" approach, by treating all costs included in his "basket" as variable. 520 Given the fact that Lundbeck AU and CNS Pharma functioned as a single budget and accounting entity, I also do not consider Mr Samuel's approach of allocation of overheads between the companies necessary. 521 These conclusions do not mean that the issue of a discount for risk in respect of the hypothetical cash flows is irrelevant. Mr Stone recognised this to be the case, despite his ultimate view that no discount need be applied. In their joint report the accountants identified two sources of risk, the first being the cost of goods sold and the second being the overheads. Lundbeck submitted this: In respect of the cost of goods sold, the potential impact of any estimation risk is relatively small. The potential impact of the cost of goods sold on the overall loss figure is modest. Further, the cost of goods sold for lost unit sales has been calculated by reference to the cost of goods actually sold in the relevant years. The cost of goods sold therefore has a sound foundation. As regards overheads, the potential for a discount relates only to the loss of Lundbeck AU and CNS Pharma. Having regard to the dispute between Mr Stone and Mr Samuel regarding to assessment of overheads, it is apparent that any risk in respect of overheads is symmetrical. That is, there is a risk of both over and understatement. In those circumstances, and given the need for the Court to adjudicate between the competing contentions relating to overheads, there is no need for a discount in respect of this aspect of the loss calculation. This is particularly so when consideration is had to the conservative assumption that has been applied to Lundbeck's lost profits claim, namely that each generic sale represents a lost sale to ESIPRAM rather than LEXAPRO. 522 Sandoz submitted that there should be an overall discount to reflect a number of matters (and, it will be recalled, I have concluded above that there must be a substantial discount of 25% to reflect the fact that some sales of Sandoz's products may have been taken from other generics rather than from Lundbeck). Apart from this issue, Sandoz referred to the risks associated with the estimation of overheads, the estimation of cost of goods sold, the accuracy of the budget, the risk inherent in hypothetical cash flows, and the relatively high discrepancy between Mr Samuel's estimation of overheads compared to Mr Stone's estimation. Of these, I accept that all but the last matter is entitled to weight. As to the last matter, I consider that the size of the discrepancy between the accounting experts is largely a result of Mr Samuel's unreasonable approach to overheads. 523 I do not accept Lundbeck's submission that all of the identified risks are off-set by the fact that the claims relate to lost sales of Esipram only and not lost saes of Lexapro (as well as reduced prices of both Esipram and Lexapro, dealt with by Ms Weber's unchallenged evidence). I agree that this assumption is conservative. Lexapro, however, was the originator's premium brand. It was more likely (and the evidence demonstrates) that Esipram was more vulnerable to generic competition than Lexapro. I nevertheless consider it highly likely that some sales made by Sandoz of its escitalopram products represented lost sales of Lexapro. The fact that the claims are founded on this conservative assumption of no lost sales of Lexapro also should be given material weight. Having regard to all of these circumstances, I consider that a further discount for risk of 5% should be deducted from Lundbeck's claims giving an overall discount of 30%.

22. Interest 524 Sandoz submitted that pre-judgment interest should not be awarded under s 51A of the Federal Court of Australia Act 1976 (Cth) in the unusual circumstances of this case. 525 I disagree. 526 As I have said, Sandoz made a calculated commercial decision to infringe the 144 patent. Before it decided to launch its generic escitalopram products Sandoz knew that Lundbeck intended to seek an extension of time to apply for an extension of term and that, if these were granted, Lundbeck alleged that Sandoz's escitalopram products would infringe the 144 patent. Sandoz decided to launch in any event knowing that, if Lundbeck succeeded, it would sue Sandoz for damages or an account of profits. Lundbeck's view that it was entitled to apply for an extension of term based on the registration of Lexapro was reasonable. So too was all of its conduct in seeking to secure its rights. It is not that Sandoz's conduct in opposing Lundbeck's applications was unreasonable. What is relevant is that Sandoz made a calculated decision to launch at risk. Its decision-makers did not launch believing that they had a licence to do so (this was a belief limited to Mr Sharkey and some others in Australia who were not the relevant decision-makers). Sandoz launched at risk. Accordingly, this is not a case, as Sandoz would have it, of Lundbeck casting "the effects of a 'self-inflicted burden'" onto Sandoz (citing H K Frost Holdings Pty Ltd (in liq) v Darvall McCutcheon (a firm) [1999] FCA 795 at [11]). In all of the circumstances, a "good cause to the contrary" of an award of interest has not been shown. 527 I note also that: (1) In State Bank of New South Wales v Commissioner of Taxation [1995] FCA 1652; (1995) 62 FCR 371 at 385 Wilcox J said: …s 51A(1) is a facultative provision intended to confer power on the Court to do justice between parties in relation to pre-judgment interest; a matter of some importance in these days of high interest rates and extensive delays in finalising litigation. The subsection should be interpreted as widely as its language allows. (2) The rate of interest is within the Court's discretion (r 39.06 of the Federal Court Rules 2011 (Cth)), but the Court's Interest on Judgments Practice Note (GPN - INT) is relevant to the discretionary exercise and informs parties that they should expect the Court to have regard to certain rates. Further, in 2009 Order 35 r 7A of the Federal Court of Australia Rules provided this: If determining a rate of interest for an order under paragraph 51A (1) (a) of the Act, the Court or a Judge may fix the rate as: (a) the cash rate of interest set by the Reserve Bank of Australia from time to time during the period mentioned in paragraph 51A (1) (a) of the Act, plus 4 per cent; or (b) such other rate as the Court or Judge thinks fit. 528 In Generic Health the Full Court said this: 263 The primary judge calculated pre-judgment interest on pre-tax loss (being the profit from lost sales). Error is said to be displayed in not assessing the tax that would have been payable on the profits that were lost, and only awarding interest on that post-tax sum. 264 It can be accepted that pre-judgment interest is compensatory in nature: Haines v Bendall [1991] HCA 15; 172 CLR 60 at 66 to 67. 265 The proper approach to that compensatory task in the context of a statutory remedy involves a choice or discretion. In Cullen v Trappell [1980] HCA 10; 146 CLR 1 at 22, Gibbs J said that the award of interest should always be approached in a broad and practical way and should not be allowed to assume disproportionate importance in the resolution of the dispute. 266 In Daniels v Anderson (1995) 37 NSWLR 438; 118 FLR 248 at 586, Clarke and Sheller JJA expressed the view that interest should generally be on a pre-tax basis. See also Hodgson v Amcor Ltd [No 9] [2012] VSC 205. 267 One of the reasons for this approach is the limited capacity of the Court, in a case about lost sales and profits, to estimate the tax liability, especially in the absence of the revenue authorities. This may, perhaps, be especially so in circumstances where one may have to attribute intra-group payments for the value of patent rights and assess legitimate intra-group costs and charges. 268 The primary judge's approach at [338] to [342] saw no call to undertake a taxation exercise. We see no error in her Honour's approach, in particular in the light of the above considerations. 529 Justice requires the award of interest given that Lundbeck has been kept out of money which it otherwise would have received for the best part of 10 years. 530 Sandoz also submitted that because the extension of term was granted on 26 June 2014 Lundbeck's cause of action for patent infringement first arose on that date so that interest may only be awarded from that date onwards. This is because s 51A refers to interest being awarded on money for the "whole or any part of the period between the date when the cause of action arose and the date as of which judgment is entered". 531 I considered s 51A in Sigma at [1282]-[1307]. I said this in particular: 1293 Similarly, the words "the date when the cause of action arose" in s 51A(1)(a) also involve a temporal requirement. If there is no period between the date on which the claim may be made for the payment of money and the date of judgment, then the provision cannot apply. 1294 In any event, in Port of Melbourne Authority v Anshun Pty Ltd (1981) 147 CLR 589 at 610 Brennan J said: There is an imprecision in the meaning of the term cause of action, which is sometimes used to mean the facts which support a right to judgment (see per Williams J. in Carter v. Egg and Egg Pulp Marketing Board (Vict.) (1942) 66 CLR 557, at pp 600, 601); sometimes to mean a right which has been infringed (see Serrao v. Noel (1885) LR 15 QBD 549), and sometimes to mean the substance of an action as distinct from its form (see Krishna Behari Roy v. Brojeswari Chowdranee (1875) LR 2 Ind App 283 ). 1295 In CGU Insurance Ltd v Watson (as trustee of the deed of arrangement in respect of Greaves) [2007] NSWCA 301 Giles JA with whom Spigelman CJ and Basten JA agreed said at [44] that what is "meant by cause of action is notoriously difficult". His Honour continued at [46]: In Onerati v Phillips Constructions Pty Ltd (in liquidation) (1989) 16 NSWLR 730, in which a question was whether a proprietor's second proceedings for breach of contract against the builder for defects discovered after judgment in the first proceedings was barred by res judicata, I said at 738-9 - "Different forms of words have been used to describe what is meant by a cause of action. Brennan J referred to three descriptions; others are every fact which it would be necessary for a plaintiff to prove, if traversed, in order to support his right to a judgment (Read v Brown (1888) 22 QBD 128 at 131 per Lord Esher); the essential ingredients in the title to the right which it is proposed to enforce (Williams v Milotin (1957) 97 CLR 465 at 474 per Dixon CJ, McTiernan, Williams, Webb and Kitto JJ; Cartledge v E Jopling& Sons Ltd [1963] AC 758 at 783-784 per Lord Pearce); the act on the part of the defendant which gives the plaintiff his cause of complaint (Jackson v Spittall (1870) LR 5 CP 542 at 552 per Brett J for himself, Bovill CJ and Keating and Montague Smith JJ; Bass v The King [1948] NZLR 777 at 781 per Gresson J; Distillers Co (Biochemicals) Ltd v Thompson [1971] AC 458 at 467 (PC); and rights which can be enforced, or liabilities which can be redressed, by legal proceedings: Sugden v Sugden [1957] P 120 at 133 per Denning LJ. I do not find minute examination of the verbal formulae particularly helpful. The form of words may vary according to the purpose for which the description is required, and in any event may not be illuminating… 532 Sandoz's submissions fail to recognise the operation of s 79 of the Patents Act which provides that once the extension is granted there is the "same rights to start proceedings in respect of the doing of an act during the period" commencing on the expiration of the term of the patent and ending on the day on which the extension was granted. This statutory fiction has effect for all purposes. It means that Lundbeck is taken to have had its rights as at the date Sandoz commenced its infringement of the 144 patent on 15 June 2009. Accordingly, in my view, the requirements of s 51A are satisfied on and from 15 June 2009. This is when the first cause of action arise. The causes of action accrued thereafter annually (for Lundbeck A/S) and on each lost sale for Lundbeck AU. I consider that it would be contrary to the terms of the legislation not to give effect to the clear intention of s 79 of the Patents Act. It follows that good cause has not been shown to confine the period of interest to the period commencing on 26 June 2014. 533 Sandoz also submitted that Mr Samuel's approach of calculating losses on a post-tax loss should be adopted in preference to Mr Stone's approach of calculating interest based on pre-tax losses. Consistent with the reasoning in Generic Health I consider that the appropriate exercise of discretion is for interest to be calculated based on pre-tax losses. The evidence of Lars Fogh, a tax lawyer in Denmark, that Lundbeck A/S would have had to pay an income tax rate of 22% for the years 2017-2019 and a corporate income tax rate of 22% on any award of damages, to my mind, is not evidence to the contrary. Whatever the tax position on the award of damages will be, tax obligations will need to be satisfied. Given the complexity of the group's tax arrangements this matter is best left to the companies themselves in the first instance and the relevant tax authorities to ascertain.

23. Misleading and deceptive conduct claim 534 Lundbeck accepted that these claims could not found any award of damages additional to the award for patent infringement if its arguments that Lundbeck A/S and Lundbeck AU were entitled to recover for the full claim period in respect of, relevantly, the lost sales of Esipram by CNS Pharma (which has no entitlement to bring a claim for patent infringement). I have accepted Lundbeck's arguments above, meaning that the claims for misleading and deceptive conduct can make no difference to the calculation of damages. 535 In these circumstances I propose to give brief reasons explaining why these alternative claims would also succeed. 536 The pleadings are to the effect that Sandoz, in supplying its escitalopram products: • failed or refused or neglected to warn its customers or potential customers including distributors, pharmacies, medical practitioners, end users and any other person or entity (customers or potential customers) that the exploitation of the Respondent's Escitalopram Products could infringe AU 144; and • impliedly represented to customers and potential customers that the customers or potential customers could use the Respondent's Escitalopram Products without infringing AU 144. 537 Lundbeck submitted this (by reference to s 18 of the Australian Consumer Law in Sch 2 to the Competition and Consumer Act 2010 (Cth), but recognising that its claims also extend back to the operation of the Trade Practices Act 1974 (Cth)): Lundbeck alleges that, by its sale and supply of generic escitalopram medicines between June 2009 and December 2012, Sandoz represented to its customers and potential customers that those customers could use those generic escitalopram medicines without the risk of infringing the patent. Lundbeck alleges that, by this conduct, Sandoz contravened s 18 of the ACL. There is no evidence that Sandoz issued any warning to its customers or potential customers about the possibility of patent infringement. However, at the time of launch, Sandoz was well aware of the risk of infringement. It has long been recognised that such a failure to warn may constitute misleading or deceptive conduct. This case is a clear example of why this is so - there was a long history of uncertainty about the expiry of the patent, but Sandoz nonetheless pursued the commercial benefits of launching without warning its customers of the risks. It appears that Sandoz might seek to distinguish the authorities referred to …above on the basis that, at the relevant time, the standard term of the patent had ended and the fate of Lundbeck's efforts to extend the term of the patent was unknown. This does not meet Lundbeck's case. When it launched ESITALO, Sandoz was well aware of the potential risk of infringing the patent. It analysed that risk at length, both through its internal legal processes and with the aid of external legal advice. It plainly considered that the risk called for careful consideration. Despite that, it did not warn its customers of the potential risk of infringement. In the circumstances, Sandoz' customers would plainly have expected to be alerted to that risk and to be given the opportunity to make a commercial decision about whether the risk was acceptable. 538 In support of these submissions Lundbeck referred to Advanced Building Systems Pty Ltd v Ramset Fasteners (Aust) Pty Ltd [1995] FCA 236; (1995) AIPC 91-129, Ramset Fasteners (Aust) Pty Ltd v Advanced Building Systems Pty Ltd [1999] FCA 898; (1999) 174 ALR 239, Sanofi-Aventis Australia Pty Ltd v Apotex Pty Ltd (No 3) [2011] FCA 846; (2011) 196 FCR 1 at [275]-[282], Apotex Pty Ltd v Sanofi-Aventis Australia Pty Ltd (No 2) [2012] FCAFC 102; (2012) 204 FCR 494 at [91] and Sandvik Intellectual Property AB v Quarry Mining & Construction Equipment Pty Ltd [2016] FCA 236; (2016) 118 IPR 421 at [272] - [277]. 539 In Advanced Building Systems [1995] FCA 236 at 38 Hill J said I am of the view that the failure of Ramset to warn its customers that use of clutches or components in a particular way might constitute an infringement of Advanced's patent, does constitute conduct in trade or commerce which is misleading or deceptive or likely to mislead or deceive users. It does not seem to me to affect the proposition to say, as counsel for Ramset says, that Ramset had no control over how, when and where components were to be used, or over its customers or their activities. So much may be conceded but does not affect the reasonable expectation that arises of a warning. Nor, in my view, does the conduct of Ramset cease to be misleading and deceptive even if the facts are, as they have been proved to be, that no legal action was ever taken against any contractors or others using clutches or components supplied by Ramset. 540 However, Hill J declined to order damages because the applicant had not proved loss caused by the failure to warn (in the sense that a customer might have decided to proceed with the purchase in any event). 541 In Ramset Fasteners [1999] FCA 898 the Full Court agreed with the conclusion that the failure to warn constituted misleading and deceptive conduct: at [66]. The Full Court said: 67. Section 52 requires that a "corporation shall not, in trade or commerce, engage in conduct that is misleading or deceptive or is likely to mislead or deceive". Attention has been drawn, in a number of cases, to the fact that the section is not limited to "representations", a word that does not appear in it, although contravening conduct is generally apt to involve representations: Henjo Investments Pty Ltd v Collins Marrickville Pty Ltd (1988) 79 ALR 83 at 93; Pacific Dunlop Ltd v Hogan (1989) 23 FCR 553 at 585-586; Demagogue Pty Ltd v Ramensky (1992) 39 FCR 31 at 32, 40-41. Whether or not Ramset's conduct should be analysed as conveying an implicit misrepresentation, it acted misleadingly when it promoted and sold face-lift tilt-up equipment in the way that it did, without informing its customers of the liability it was inducing them to incur. This was conduct calculated to cause a mistaken impression about a significant consequence of the transaction proposed to those customers. In our opinion, in such a case, it is unnecessary, and may be artificial, to speak of a representation. The orthodox theory would find one, but it is more realistic to see the conduct as misleading, without resorting to a sophisticated analysis in which no one would have engaged at the time. 68. It is plain, on the evidence and the findings of the trial judge, that Ramset's conduct was engaged in for the purpose of persuading customers to purchase the various forms of equipment it supplied for face-lift tilt-up operations. Many persons did so. For example, Mr Nightingale gave evidence, to which reference has been made, of a big change in the source of supply upon which Brambles Cranes drew to obtain ring clutches. In Como Investments Pty Ltd v Yenald Nominees Pty Ltd (1997) ATPR 43,617 at 43,619-43,620, a full court jointly stated: "Where a representation is relevant to the decision in question, and in its nature persuasive to induce the making of that decision, it accords with legal notions of causation to hold that it has a causative effect. And where a respondent, who may be taken to know his own business, has thought it was in his interests to misrepresent the situation in a particular respect, the Court may infer that the misrepresentation was persuasive. These inferences arise from the making of the representation followed by the respondent doing the thing it was calculated to induce him to do." Those observations may be applied equally to the effect of the misleading conduct of Ramset upon its customers. See also Gould v Vaggelas (1985) 157 CLR 215 at 236; Commission for the New Towns v Cooper (Great Britain) Ltd [1995] Ch 259 at 282; Sibley v Grosvenor (1916) 21 CLR 469 at 473, 478, 481; Krakowski v Eurolynx Properties Limited (1995) 183 CLR 563 at 578; Hanave Pty Ltd v LFOT Pty Ltd (1999) FCA 357 at paras 37, 45. In the present case, the inference should be drawn that Ramset's misleading conduct did cause damage to Advanced, but it is unnecessary to pursue this point further. 542 In Sanofi-Aventis [2011] FCA 846 at [281] I accepted that the supply of a drug infringing a patent involved misrepresentations conveyed to medical practitioners, pharmacists and patients of the kind alleged by Lundbeck in the present case. 543 In the appeal judgment in Sanofi-Aventis [2012] FCAFC 102 at [91] the Full Court said that the misrepresentation issue depended on the existence of the patent infringement. 544 In Sandvik at [276] Jessup J said: 276. Failure to warn may amount to misleading conduct under s 52 of the TP Act and s 18 of the ACL: Winterton Constructions Pty Ltd v Hambros (1992) 39 FCR 97, 114. In the context of a patent case, in Ramset the Full Court said (164 ALR at 268 [67]): … Whether or not Ramset's conduct should be analysed as conveying an implicit misrepresentation, it acted misleadingly when it promoted and sold face-lift tilt-up equipment in the way that it did, without informing its customers of the liability it was inducing them to incur. This was conduct calculated to cause a mistaken impression about a significant consequence of the transaction proposed to those customers. In our opinion, in such a case, it is unnecessary, and may be artificial, to speak of a representation. The orthodox theory would find one, but it is more realistic to see the conduct as misleading, without resorting to a sophisticated analysis in which no one would have engaged at the time. 545 Sandoz's submissions to the contrary did not confront these decisions. It is not necessary to go so far as to say that there is a general principle that a failure to warn of a potential patent infringement by the supply of a product is likely to involve misleading and deceptive conduct. It may be accepted that each case turns on its own facts. The problem for Sandoz is that the facts of the present case provide no basis upon which to distinguish the approach in Sanofi-Aventis. Sandoz again ignored the effect of s 79 of the Patents Act when it submitted that in cases in which misleading and deceptive conduct had been found in the context of a failure to warn of patent infringement the patents were in force and effect. Given the terms of s 79 and the facts of the present case, this is not a relevant distinguishing feature. In my view, there is no relevant distinguishing feature. Sandoz achieved its sales to pharmacists on the basis of an implied misrepresentation that its products did not infringe any patent, but the products would infringe if and when the extension of term was granted. The extension of term was granted and thus the products infringed. This is sufficient to constitute misleading and deceptive conduct in the circumstances. No additional damages may be recovered given my conclusions but, consistent with the authorities, if I had reached any different view I do not see why damages would not be recoverable. 546 To the extent necessary to say so I would also accept Lundbeck's submissions as follows: It appears that Sandoz will also contend that the claims pursued by Lundbeck DK and Lundbeck AU under the ACL in proceeding NSD 647 of 2014 are statute barred in respect of events which occurred prior to 10 December 2010. The basis for Sandoz' contention is that Lundbeck was granted leave to make the amendments introducing these claims on 9 December 2016. The claims of Lundbeck DK and Lundbeck AU should not be limited in this manner and the amendments to introduce these claims should be given effect from the date of the commencement of proceeding NSD 647 of 2014 (ie 26 June 2014). These claims arise by reason of substantially the same facts as are relied upon to support the patent infringement claim. While it may be accepted that the case under the ACL also involves Sandoz having failed to warn its customers of the possibility of infringement, that matter is uncontroversial and is merely an incident of Sandoz' allegedly infringing conduct. The Court has the power to treat the amendments as effective from the commencement of the proceeding, irrespective of the timing of the amendments themselves: see Federal Court of Australia Act 1976 (Cth), s 59(2B) and Federal Court Rules, rule 8.21(2). It is in the interests of justice for the Court to do so. Lundbeck DK and Lundbeck AU have been pursuing their rights in respect of the relevant conduct of Sandoz since they commenced proceeding NSD 647 of 2014. The amendments merely approach that conduct in a different manner. In those circumstances, it would be unjust to limit the claims of Lundbeck DK and Lundbeck AU in the manner apparently proposed by Sandoz. 547 I do not accept Sandoz's submissions that Lundbeck's damages (if any were payable which, given my conclusions about damages patent infringement, they could not be) should be reduced. Sandoz said: If the Court were to find Sandoz liable for contravention of the ACL (or TPA) in respect of any part of the relevant period, the amount of damages awarded ought to be reduced significantly having regard to the Applicants' own responsibility for any relevant loss or damage they suffered, pursuant to CCA, s 137B; TPA, s 82(1B). Section 137B of the CCA provides that where a person (claimant) makes a claim under s 236(1) of the ACL in relation to economic loss, suffered by the claimant because of the conduct of another person, and the conduct contravened s 18 of the ACL, and the claimant suffered loss and damage as a result, partly because of the claimant's failure to take reasonable care, and party of the conduct of the other person, and the other person did not intend to cause the loss or damage (and did not fraudulently cause the loss or damage), the amount of loss of damage that the claimant may recover is to be reduced to the extent to which the Court thinks just and equitable having regard to the claimant's share in responsibility for the loss or damage. 548 However, as Lundbeck submitted, and consistent with my conclusions above about Lundbeck's conduct being reasonable and Sandoz having supplied its products at risk based on its own commercial calculations, Sandoz's contentions in this regard are without merit as: Lundbeck did not encourage or countenance the impugned conduct of Sandoz. To the contrary, Sandoz knowingly launched at risk of infringement allegations being made by Lundbeck. In due course, Lundbeck pursued its infringement case at the earliest opportunity.

24. Conclusions 549 Lundbeck should succeed in its claims for damages for patent infringement generally in accordance with the approach taken by Mr Stone, subject to an overall discount of 30% to take account of the risks which I have identified in these reasons for judgment. I am aware that Mr Stone's further calculations provided to the Court by email on 13 November 2018 have not yet been reviewed by Mr Samuel. Further orders will provide this opportunity for review to Mr Samuel and for both Mr Stone and Mr Samuel to inform me of any outstanding issues about the calculations provided that these reasons for judgment are used as the basis for them. 550 I am also aware that Sandoz has pleaded that if it required a licence from Lundbeck to supply its escitalopram products (which, on my conclusions, it did from 15 June 2009 until 9 December 2012 and did not have a licence (which I have concluded it did not), then it has applied for such a licence under s 223(9) of the Patents Act. I have been informed that the application for a licence will be heard on 19 and 20 November 2018. My preliminary view is that the appropriate way to deal with this is for me to make final orders as appropriate once all required calculations are complete and for me to hear and determine any application for a stay of those orders as appropriate. 551 Before general publication of these reasons for judgment, the parties should also be given an opportunity to consider these reasons for the purpose of deciding if any claim for confidentiality should be made and proposing further directions to bring this long-running litigation to an end as soon as possible. I certify that the preceding five hundred and fifty-one (551) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Jagot.