Bristol-Myers Squibb Company v Apotex Pty Ltd - [2013] FCA 1114 - FCA 2013 case summary — Zoe

CTHFCA

Bristol-Myers Squibb Company v Apotex Pty Ltd

[2013] FCA 1114

Federal Court of Australia|2013-10-30|Before: Yates J

View original source

At a glance

Source facts

Court

Federal Court of Australia

Decision date

2013-10-30

Before

Yates J

Source

Original judgment source is linked above.

Judgment (41 paragraphs)

[1]

Introduction 1 The applicants allege that the respondent, Apotex Pty Ltd (Apotex), has infringed or threatened to infringe certain claims of Patent No. 2002334413, which is entitled "Low hygroscopic aripiprazole drug substance and processes for the preparation thereof" (the patent). Aripiprazole is an atypical antipsychotic agent that is useful for the treatment of schizophrenia. 2 The second applicant, Otsuka Pharmaceutical Co., Ltd (Otsuka), is the patentee. The first applicant, Bristol-Myers Squibb Company (BMS), claims to be the exclusive licensee of the patent. 3 Prior to 20 October 2009, Apotex offered to sell in Australia variously-branded pharmaceutical products that contain aripiprazole as the active ingredient (the Apotex products). The Apotex products are included in the Australian Register of Therapeutic Goods (the ARTG) for the treatment of schizophrenia. In essence, the applicants alleged that Apotex's sale of the Apotex products infringes or would lead to infringement of the patent. On 20 October 2009, Apotex agreed to orders that restrained it from supplying the Apotex products until the applicants' claims were determined. 4 Apotex challenged the validity of the claims asserted against it. It contended that the invention as claimed is not novel, does not involve an inventive step, and is not a manner of manufacture within the meaning of s 6 of the Statute of Monopolies. It also contended that some of the claims are not clear or fairly based on the matter described in the complete specification of the patent. Further, it contended that the patent was obtained on a false suggestion or misrepresentation. In these reasons, I deal with the grounds of invalidity advanced in the order in which the parties addressed them in their submissions, namely, novelty, false suggestion, inventive step, manner of manufacture, clarity, and fair basis. 5 The present hearing is concerned only with the validity of the asserted claims and Apotex's liability for infringement. These questions are governed by the Patents Act 1990 (Cth) (the Act) in the form in which it existed as at 1 April 2002, taking into account the amendments made by Act No. 160 of 2001. 6 For the reasons that follow, I have concluded that Apotex's challenges to the validity of the asserted claims fail, other than in respect of its challenge to claim 45 which, in my view, does not claim an invention that is a manner of manufacture within the meaning of the Statute of Monopolies. I have also concluded that the second applicant's case on infringement succeeds on all asserted claims other than in respect of claim 45 (which is invalid), claim 43 which is directed to the preparation of a medicament to treat schizophrenia and its symptoms, and claim 14, in respect of which I wish to receive further submissions.

[8]

Powder x-ray diffraction 22 The shape of the unit cell is defined by three unit cell lengths and three unit cell angles (the lattice parameters). The unit cell has a definite volume that contains the atoms, molecules, and ions necessary for generating the crystal. 23 X-ray diffraction (XRD) is an analytical technique that is used to obtain information about the unit cell shape and structure of a crystalline substance. It is based on the diffraction of x-rays by crystalline solids. In this technique, a beam of x-rays, typically of the same wavelength, is directed at a crystalline substance. When the beam hits the substance, it is diffracted by the atoms, molecules or ions in the crystal structure at certain special orientations, creating a diffractogram or diffraction pattern (XRD pattern). The lattice parameters determine these special orientations and, hence, the positions of peaks shown in the XRD pattern. 24 Powder x-ray diffraction (XRPD or, sometimes, PXRD) is a particular form of XRD in which the sample of interest is crushed or ground to a fine powder and analysed using an x-ray diffractometer. The x-rays are confined to a parallel (collimated) beam and directed onto the sample, which is located at the centre of the instrument. The angles through which the incident x-ray beams are diffracted are measured by a moveable detector arm. The angle measured by the detector through which the incident x-ray beam is diffracted is twice the angle θ (theta). The angle is known as °2θ. It is common for XRPD results to be quoted by reference to this angle. 25 In the powder sample, there are a multitude of small single crystals or "crystallites". The object is for the crystallites to be packed into or adhered to a sample holder in random orientations. The x-ray source, sample, and detector can then be moved through a range of angles to measure the diffraction pattern generated by the powder (the XRPD pattern). The results obtained are a series of diffraction angles and x-ray intensities measured as a function of the angle °2θ. These results are recorded by a computer. Software is then used to generate an XRPD pattern that may be used to pick the peak positions. The locations of the peaks in the XRPD pattern are related to the spacing between the planes in the crystal lattice according to Bragg's law: nλ = 2d sin θ. 26 Importantly, every crystalline form of a particular crystalline substance has a unique unit cell that produces a unique XRPD pattern. This pattern is like a fingerprint. It distinguishes one crystalline form of that substance from other crystalline forms. For any given crystalline form, the positions of the diffracted peaks are precisely determined by Bragg's law. 27 The XRPD pattern obtained for any given sample is subject to variation depending on a number of factors. When comparing sets of XRPD data, it is important to ensure that the parameters or settings (notably, the wavelength of the x-ray source) are the same (or that any differences are taken into account) as the measured angles of diffraction will vary.

[12]

The witnesses 37 The applicants adduced evidence from the following witnesses: Satoshi Aoki. Mr Aoki is named as an inventor in the patent. He is an associate manager and researcher in the Bulk Pharmaceutical Chemicals Department at Otsuka's Second Tokushima Factory. Mr Aoki made one affidavit that was read at the hearing. Clive Allan Prestidge. Professor Prestidge is a professor of colloid and pharmaceutical science at the Ian Wark Research Institute, University of South Australia. Professor Prestidge made one affidavit that was read at the hearing. Christopher John Easton. Professor Easton is a professor at the Research School of Chemistry, Institute of Advanced Studies, Australian National University. Professor Easton made two affidavits that were read at the hearing. Jonathan Michael White. Associate Professor White is an associate professor and reader at the School of Chemistry, University of Melbourne. He made one affidavit that was read at the hearing. Raymond Leslie Withers. Professor Withers is a professor and head of the Materials Chemistry Group at the Research School of Chemistry, Australian National University. He made one affidavit that was read at the hearing. 38 Apotex adduced evidence from the following witnesses: Ross Peter McGeary. Associate Professor McGeary is an associate professor with a shared appointment in the School of Chemistry and Molecular Biosciences and the School of Pharmacy at the University of Queensland. He made two affidavits that were read at the hearing. David St Clair Black. Professor Black is a professor of organic chemistry at the University of New South Wales. He made two affidavits that were read at the hearing. James Steven Rowe. Dr Rowe is a pharmaceutical formulation chemist and consultant. He made two affidavits that were read at the hearing. 39 These witnesses, other than Professor Withers, were cross-examined. Despite some criticisms by the parties in final submissions of the evidence given by some witnesses, I found the evidence of all witnesses to be helpful in understanding the scientific issues arising in this case. Where criticisms have been made, I have dealt with them, to the extent that I consider necessary, when discussing aspects of the evidence later in these reasons.

[13]

The patent 40 The patent is a standard patent. It was granted on a PCT application (PCT/JP02/09858) taken as filed on 25 September 2002 and published as WO 03/026659. The claims of the patent claim priority from three basic applications: Japanese Patent Application No. 2001-290645 filed on 25 September 2001; Japanese Patent Application No. 2001-348276 filed on 14 November 2001; and Canadian Patent Application No. 2379005 filed on 27 March 2002. Thus, the earliest priority date of the claims of the patent is 25 September 2001. 41 The patent claims were amended by an order made pursuant to s 105 of the Act on 13 August 2010: see Bristol-Myers Squibb Co and Another v Apotex Pty Ltd (ACN 096 916 148) (2010) 87 IPR 516 (the amendment reasons). 42 The invention described in the patent relates to an improved form of aripiprazole having reduced hygroscopicity. 43 The complete specification describes, as background to the invention, two published methods of manufacturing anhydrous crystals of aripiprazole. The term "anhydrous" means that the compound contains no combined water. 44 The first method is disclosed in Example 1 of Japanese Unexamined Patent Publication No. 191256/1990 (the Japanese unexamined patent publication). It involves reacting 7-(4-bromobutoxy)-3,4-dihydrocarbostyril with 1-(2,3-dichlorophenyl)piperazine to produce raw anhydrous aripiprazole, which is then recrystallised with ethanol. The second method is disclosed in Proceedings of the 4th Japanese-Korean Symposium on Separation Technology (6-8 October 1996) (the Symposium), specifically, in documents referred to in the present case as the Aoki article and the Aoki poster. This method involves heating aripiprazole hydrate at 80°C. 45 The disclosures of the 141 application, the Aoki article, and the Aoki poster, which are described in greater detail below, constitute important prior art information. They are the foundation for Apotex's case that the invention, as relevantly claimed, is not novel and is obvious, and that the patent was obtained on a false suggestion. 46 The complete specification states that the anhydrous crystals obtained by the two disclosed methods have the disadvantage of being significantly hygroscopic. The complete specification describes this disadvantage as follows: The hygroscopicity of these crystals makes them difficult to handle since costly and burdensome measures must be taken in order to ensure they are not exposed to moisture during process and formulation. Exposed to moisture, the anhydrous form can take on water and convert to a hydrous form. This presents several disadvantages. First, the hydrous forms of aripiprazole have the disadvantage of being less bioavailable and less dissoluble than the anhydrous forms of aripiprazole. Second, the variation in the amount of hydrous versus anhydrous aripiprazole drug substance from batch to batch could fail to meet specifications set by drug regulatory agencies. Third, the milling may cause the drug substance, Conventional Anhydrous Aripiprazole, to adhere to manufacturing equipment which may further result in processing delay, increased operator involvement, increased cost, increased maintenance and lower production yield. Fourth, in addition to problems caused by introduction of moisture during the processing of these hygroscopic anhydrous crystals, the potential for absorbance of moisture during storage and handling would adversely affect the dissolubility of aripiprazole drug substance. Thus shelf-life of the product could be significantly decreased and/or packaging costs could be significantly increased. It would be highly desirable to discover a form of aripiprazole that possessed low hygroscopicity thereby facilitating pharmaceutical processing and formulation operations required for producing dosage units of an aripiprazole medicinal product having improved shelf-life, suitable dissolubility and suitable bioavailability. 47 The complete specification says that these anhydrous crystals (which are referred to in other passages of the complete specification as "conventional" anhydrous aripiprazole crystals) (conventional anhydrous aripiprazole crystals) exist as "type-I" crystals and "type-II" crystals. Type-I crystals can be prepared by the methods discussed above. Type-II crystals can be prepared by heating the type-I crystals at 130°C to 140°C for 15 hours. However, by using these methods, type-II crystals having high purity cannot be easily prepared on an industrial scale with good repeatability. 48 The type-I and type-II crystals are referred to in the Aoki article as type 1 and type 2 crystals, respectively. I will use that nomenclature when further discussing the complete specification and the evidence. As I will explain later, the Aoki article also refers to type 3 crystals in the form of a hydrate. 49 The complete specification summarises the invention in the following terms: Thus according to the present invention there is provided a form of aripiprazole having reduced hygroscopicity and which is more amenable to pharmaceutical processing and formulation. The inventors of the present invention have discovered that this reduced-hygroscopic form of Aripiprazole is a crystalline substance defined herein as Anhydrous Aripiprazole Crystals B. A particular process for the preparation of this anhydrous crystalline substance has also been discovered and comprises yet another aspect of the present invention. Particularly, it was discovered as part of the present invention that in order to produce Anhydrous Aripiprazole Crystals B having the desired pharmaceutical properties and utilizing the most efficient process, Hydrate A, as defined herein, would have to serve as the intermediate. It was also discovered that a particular sequence of processing had to be implemented in order to form Hydrate A. It was discovered that the preparation of Hydrate A required milling what is defined herein as Conventional Hydrate. Then, Hydrate A can be transformed into Anhydrous Aripiprazole Crystals B through suitable heating as defined herein. Surprisingly, if the Conventional Hydrate is first heated and then milled, serious agglomeration sets in rendering the processing commercially unsuitable. Advantageously at least one embodiment of the present invention provides novel anhydrous aripiprazole crystals. Moreover, another advantage of at least one embodiment of the present invention is that anhydrous aripiprazole crystals which neither easily convert into hydrates nor substantially decrease the original solubility, even when a pharmaceutical composition comprising anhydrous aripiprazole is stored for a long period of time may be provided. A further advantage of at least one embodiment of the present invention is that preparation methods, in order to obtain anhydrous aripiprazole crystals having high purity in an industrial scale with good repeatability may be provided. The present inventors have conducted research works aimed to attain the aforementioned advantages. In the course of the research, they have found that the desired anhydrous aripiprazole crystals can be obtained when a well-known anhydrous aripiprazole is heated at a specific temperature. Further, the present inventors have found that the desired anhydrous aripiprazole crystals can be obtained from recrystallization of a well-known anhydrous aripiprazole by using the specific solvents. Moreover, the present inventors found that the desired anhydrous aripiprazole crystals can be obtained by suspending a well-known anhydrous aripiprazole in the specific solvent, and heating the thus obtained suspension. The present invention is thus completed on the basis of these findings and knowledge. 50 To summarise, these passages from the complete specification disclose a form of aripiprazole that is more amenable to pharmaceutical processing and formulation as well as the most efficient process then known by which that form can be obtained. That process uses Hydrate A as an intermediate which is then transformed into Anhydrous Aripiprazole Crystals B (Crystals B) "through suitable heating". 51 Despite what is said in the complete specification, Conventional Hydrate (conventional hydrate) is not defined. However, a process for preparing conventional hydrate is described. The complete specification states that anhydrous aripiprazole crystals obtained from the method in Example 1 of the Japanese unexamined patent publication are dissolved in a hydrous solvent, heated, and then cooled, such that aripiprazole hydrate is precipitated as crystals in the hydrous solvent. The complete specification describes the hydrous solvent in the following terms: An organic solvent containing water is usually used as the hydrous solvent. The organic solvent should be one which is miscible with water, such as for example an alcohol such as methanol, ethanol, propanol or isopropanol, a ketone such as acetone, an ether such as tetrahydrofuran, dimethylformamide, or a mixture thereof, with ethanol being particularly desirable. The amount of water in the hydrous solvent can be 10-25% by volume of the solvent, or preferably close to 20% by volume. 52 The complete specification discloses a process for manufacturing Hydrate A: Hydrate A is manufactured by milling Conventional Hydrate. Conventional milling methods can be used to mill Conventional Hydrate. For example, Conventional Hydrate can be milled in a milling machine. A widely used milling machine can be used, such as an atomizer, pin mill, jet mill or ball mill. Of these, the atomizer is preferred. Regarding the specific milling conditions when using an atomizer, a rotational speed of 5000-15000 rpm could be used for the main axis, for example, with a feed rotation of 10-30 rpm and a screen hole size of 1-5 mm. The mean particle size of the Aripiprazole Hydrate A obtained by milling should normally be 50 µm or less, preferably 30 µm or less. Mean particle size can be ascertained by the particle size measurement method described hereinafter. 53 The complete specification characterises Hydrate A in the following terms: Particles of "Hydrate A" as used herein have the physicochemical properties given in (1) - (5) below: (1) It has an endothermic curve which is substantially the same as the thermogravimetric/differential thermal analysis (heating rate 5°C/min) endothermic curve shown in Figure 1. Specifically, it is characterized by the appearance of a small peak at about 71°C and a gradual endothermic peak around 60°C to 120°C. (2) It has an 1H-NMR spectrum which is substantially the same as the 1H-NMR spectrum (DMSO-d6, TMS) shown in Figure 2. Specifically, it has characteristic peaks at 1.55-1.63 ppm (m, 2H), 1.68-1.78 ppm (m, 2H), 2.35-2.46 ppm (m, 4H), 2.48-2.56 ppm (m, 4H + DMSO), 2.78 ppm (t, J = 7.4 Hz, 2H), 2.97 ppm (brt, J = 4.6 Hz, 4H), 3.92 ppm (t, J = 6.3 Hz, 2H), 6.43 ppm (d, J = 2.4 Hz, 1H), 6.49 ppm (dd, J = 8.4 Hz, J = 2.4 Hz, 1H), 7.04 ppm (d, J = 8.1 Hz, 1H), 7.11-7.17 ppm (m, 1H), 7.28-7.32 ppm (m, 2H) and 10.00 ppm (s, 1H). (3) It has a powder x-ray diffraction spectrum which is substantially the same as the powder x-ray diffraction spectrum shown in Figure 3. Specifically, it has characteristic peaks at 20 = 12.6°, 15.4°, 17.3°, 18.0°, 18.6°, 22.5° and 24.8°. (4) It has clear infrared absorption bands at 2951, 2822, 1692, 1577, 1447, 1378, 1187, 963 and 784 cm-1 on the IR (KBr) spectrum. (5) It has a mean particle size of 50 µm or less. 54 The complete specification characterises Crystals B in the following terms: "Anhydrous Aripiprazole Crystals B" of the present invention as used herein have the physicochemical properties given in (6) - (12) below. (6) They have an 1H-NMR spectrum which is substantially the same as the 1H-NMR spectrum (DMSO-d6, TMS) shown in Figure 4. Specifically, they have characteristic peaks at 1.55-1.63 ppm (m, 2H), 1.68-1.78 ppm (m, 2H), 2.35-2.46 ppm (m, 4H), 2.48-2.56 ppm (m, 4H + DMSO), 2.78 ppm (t, J = 7.4 Hz, 2H), 2.97 ppm (brt, J = 4.6 Hz, 4H), 3.92 ppm (t, J = 6.3 Hz, 2H), 6.43 ppm (d, J = 2.4 Hz, 1H), 6.49 ppm (dd, J = 8.4 Hz, J = 2.4 Hz, 1H), 7.04 ppm (d, J = 8.1 Hz, 1H), 7.11-7.17 ppm (m, 1H), 7.28-7.32 ppm (m, 2H) and 10.00 ppm (s, 1H). (7) They have a powder x-ray diffraction spectrum which is substantially the same as the powder x-ray diffraction spectrum shown in Figure 5. Specifically, they have characteristic peaks at 20 = 11.0°, 16.6°, 19.3°, 20.3° and 22.1°. (8) They have clear infrared absorption bands at 2945, 2812, 1678, 1627, 1448, 1377, 1173, 960 and 779 cm-1 on the IR (KBr) spectrum. (9) They exhibit an endothermic peak near about 141.5°C in thermogravimetric/differential thermal analysis (heating rate 5°C/min). (10) They exhibit an endothermic peak near about 140.7°C in differential scanning calorimetry (heating rate 5°C/min). (11) Anhydrous Aripiprazole Crystals B of the present invention have low hygroscopicity. For example, Anhydrous Aripiprazole Crystals B of the present invention maintain a water content of 0.4% or less after 24 hours inside a closed container set at a temperature of 60°C and a humidity of 100%. Well-known methods of measuring water content can be used as long as they are methods commonly used for measuring the water content of crystals. For example, a method such as the Karl Fischer method can be used. (12) When the small particle size is required for the formulation such as tablet and other solid dose formulations including for example flashmelt formulations, the mean particle size is preferably 50 µm or less. 55 The complete specification discloses the following processes by which Crystals B can be manufactured: In case of the formulation for which small particle size (less than 50 µm) is required, the milling is necessary for the preparation. However, when a large amount of Conventional Anhydrous Aripiprazole or Anhydrous Crystals B having large particle size is milled, the milled substances adhere with each other in the milling machine. Accordingly, there is a disadvantage that it is difficult to industrially prepare Anhydrous Aripiprazole Crystals B having small particle size. Under the circumstances, the inventors of the present invention have found that Conventional Hydrate can be easily milled, and Anhydrous Aripiprazole Crystals B having small particle size can be obtained in high yield with good-operability by heating the milled hydrate A thus obtained. The Anhydrous Aripiprazole Crystals B of the present invention are prepared for example by heating the aforementioned Aripiprazole Hydrate A at 90-125°C. The heating time is generally about 3-50 hours, but cannot be stated unconditionally since it differs depending on heating temperature. The heating time and heating temperature are inversely related, so that for example the heating time will be longer the lower the heating temperature, and shorter the higher the heating temperature. Specifically, if the heating temperature of Aripiprazole Hydrate A is 100°C, the heating time should normally be 18 hours or more or preferably about 24 hours. If the heating temperature of Aripiprazole Hydrate A is 120°C, on the other hand, the heating time can be about 3 hours. The Anhydrous Aripiprazole Crystals B of the present invention can be prepared with certainty by heating Aripiprazole Hydrate A for about 18 hours at 100°C, and then heating it for about 3 hours at 120°C. The Anhydrous Aripiprazole Crystals B of the present invention can also be obtained if the heating time is extended still further, but this may not be economical. When small particle size is not required for the formulation, e.g., when drug substance is being manufactured for injectable or oral solution formulations, Anhydrous Aripiprazole Crystals B can be also obtained [by] the following process. The inventors also discovered that it is possible to obtain anhydrous aripiprazole crystals by heating conventional aripiprazole hydrate or conventional anhydrous aripiprazole crystals to a specific temperature but this process does not yield Anhydrous Aripiprazole Crystals B as a crystalline substance suitable for commercial use in the formulation of solid oral dose formulations. Furthermore, the Anhydrous Aripiprazole Crystals B of the present invention are prepared for example by heating conventional anhydrous aripiprazole crystals at 90-125°C. The heating time is generally about 3-50 hours, but cannot be stated unconditionally since it differs depending on heating temperature. The heating time and heating temperature are inversely related, so that for example the heating time will be longer the lower the heating temperature, and shorter the higher the heating temperature. Specifically, if the heating temperature of the anhydrous aripiprazole crystals is 100°C, the heating time can be about 4 hours, and if the heating temperature is 120°C the heating time can be about 3 hours. 56 Later, the complete specification discloses processes for preparing conventional (or crude) aripiprazole crystals, conventional anhydrous aripiprazole, and conventional hydrate. It again discloses a process for preparing Crystals B. 57 In relation to the preparation of conventional (or crude) aripiprazole crystals, the complete specification exemplifies, once again, Example 1 of the Japanese unexamined patent publication (although, in more detail). In relation to the preparation of conventional anhydrous aripiprazole crystals, the complete specification merely states that the method is that described in the Symposium (in other words, the Aoki article and the Aoki poster). Despite the different nomenclature used in this part of the complete specification, both described methods result in the production of conventional anhydrous aripiprazole crystals. 58 In relation to the preparation of Crystals B, the complete specification exemplifies the heating of conventional hydrate at 90°C to 125°C in terms closely similar to the penultimate paragraph in the passage quoted in [55] above which concerns the heating of conventional anhydrous aripiprazole crystals. 59 The complete specification provides a number of reference examples. It is necessary to refer to three of them. 60 Reference Example 1 is in the following terms: 19.4 g of 7-(4-chlorobutoxy)-3,4-dihydrocarbostyril and 16.2 g of 1-(2,3-dichlorophenyl) piperadine [sic] 1 hydrochloride were added to 8.39 g of potassium carbonate dissolved in 140 ml of water, and circulated for 3 hours under agitation. After reaction the mixture was cooled and the precipitated crystals filtered out. These crystals were dissolved in 350 ml of ethyl acetate, and about 210 ml of water/ethyl acetate azeotrope removed under reflux. The remaining solution was cooled, and the precipitated crystals filtered out. The resulting crystals were dried for 14 hours at 60°C to produce 20.4 g (74.2%) of raw aripiprazole. 30 g of the raw aripiprazole obtained above was recrystallized from 450 ml of ethanol according to the methods described in Japanese Unexamined Patent Publication No. 191256/1990, and the resulting crystals dried for 40 hours at 80°C to obtain anhydrous aripiprazole crystals. The yield was 29.4 g (98.0%). The melting point (mp) of these anhydrous aripiprazole crystals was 140°C, matching the melting point of the anhydrous aripiprazole crystals described in Japanese Unexamined Patent Publication No. 191256/1990. When these crystals were left for 24 hours in a dessicator set at humidity 100%, temperature 60°C, they exhibited hygroscopicity of 3.28% … 61 Reference Example 2 is in the following terms: 6930 g of the intermediate raw aripiprazole obtained in Reference Example 1 was heat dissolved in 138 liters of hydrous ethanol (water content 20%) according to the method presented at the 4th Japanese-Korean Symposium on Separation Technology, gradually (2-3 hours) cooled to room temperature, and then chilled to near 0°C. The precipitated crystals were filtered out, producing about 7200 g of aripiprazole hydrate (wet state). The wet-state aripiprazole hydrate crystals obtained above were dried for 30 hours at 80°C to obtain 6480 g (93.5%) of conventional anhydrous aripiprazole crystals. The melting point (mp) of these crystals was 139.5°C. These crystals were confirmed by the Karl Fischer method to be anhydrous, with a moisture value of 0.03%. When left for 24 hours in a closed container set at humidity 100%, temperature 60°C, these crystals exhibited hygroscopicity of 1.78% … 62 Mr Aoki gave evidence that the crystals referred to in Reference Example 2 were type 1 crystals, as referred to in the Aoki article (corresponding to type-I crystals referred to in the complete specification). 63 Reference Example 3 is in the following terms: 820 g of the intermediate wet-state aripiprazole hydrate obtained in Reference Example 2 was dried for 2 hours at 50°C to obtain 780 g of aripiprazole hydrate crystals. These crystals had a moisture value of 3.82% according to the Karl Fischer method. As shown in Figure 6, thermogravimetric/differential thermal analysis revealed endothermic peaks at 75.0, 123.5 and 140.5°C. Because dehydration began near 70°C, there was no clear melting point (mp). As shown in Figure 7, the powder x-ray diffraction spectrum of aripiprazole hydrate obtained by this method exhibited characteristic peaks at 20 = 12.6°, 15.1°, 17.4°, 18.2°, 18.7°, 24.8° and 27.5°. The powder x-ray diffraction spectrum of this aripiprazole hydrate was identical to the powder x-ray diffraction spectrum of aripiprazole hydrate presented at the 4th Joint Japanese-Korean Symposium on Isolation Technology. 64 It is clear from the last paragraph of this description in Reference Example 3 that the hydrate so produced corresponds to the type 3 crystals referred to in the Aoki article. 65 After providing the reference examples, the complete specification provides a number of further examples. Examples 1 to 10 are relevant to the present case. 66 Example 1 concerns the production of Hydrate A by milling the aripiprazole hydrate crystals obtained in Reference Example 3. 67 Example 2 concerns the production of Crystals B by drying Hydrate A for 24 hours at 100°C. Examples 3 and 4 also concern the production of Crystals B by heating Hydrate A. In each case, the steps involved the hot-air drying of Hydrate A for 18 hours at 100°C and then heating it for three hours at 120°C. 68 Examples 5 to 10 concern the production of Crystals B by heating conventional anhydrous aripiprazole crystals or conventional hydrate. In each case, the heating was carried out at 100°C (Examples 5, 7, and 9) or at 120°C (Examples 6, 8, and 10) on conventional anhydrous aripiprazole crystals obtained from the method in Reference Example 1 (Examples 5 and 6) and Reference Example 2 (Examples 7 and 8) or on conventional hydrate obtained from the method in Reference Example 3 (Examples 9 and 10). The complete specification says that Examples 5 to 10 are useful for injectable or oral solution formulations but not solid dose formulations because they were made by heating conventional anhydrous crystals or conventional hydrate instead of Hydrate A. 69 In each of Examples 2 to 10, the resulting crystals (that is, Crystals B) were left for 24 hours in a dessicator or closed container set at 100% humidity and 60°C. In each case, the crystals did not exhibit hygroscopicity exceeding 0.40%. Indeed, in each case, the hygroscopicity of these crystals was well below that percentage figure. However, when the resulting crystals of Reference Examples 1 and 2 were tested in the same way, the crystals, in each case, exhibited hygroscopicity well in excess of 0.40%. In the case of Reference Example 1, the hygroscopicity of the crystals was 3.28% and, in the case of Reference Example 2, the hygroscopicity was 1.78%. 70 The complete specification describes a number of other anhydrous aripiprazole crystals, designated respectively as Anhydrous Aripiprazole Crystals C to G. These crystals differ in their characteristics and are not directly relevant to the present case. 71 Claim 12 is of central significance to the case. Claim 12 claims Crystals B as follows: Anhydrous Aripiprazole Crystals B wherein said crystals have low hygroscopicity wherein said low hygroscopicity is a moisture content of 0.40% or less after placing said drug substance for 24 hours in a closed container maintained at a temperature of 60°C and a humidity level of 100%; have a powder x-ray diffraction spectrum which is substantially the same as the following powder x-ray diffraction spectrum shown in Figure 5; have particular infrared absorption bands at 2945, 2812, 1678, 1627, 1448, 1377, 1173, 960 and 779 cm-1 on the IR (KBr) spectrum; exhibit an endothermic peak near about 141.5°C in thermogravimetric/differential thermal analysis (heating rate 5°C/min); and exhibit an endothermic peak near about 140.7°C in differential scanning calorimetry (heating rate 5°C/min). 72 A number of other claims are dependent, directly or indirectly, on claim 12 and are relied on by the applicants for their case on infringement. Those claims are claims 13, 14, 16, 30, 31, 35, 36, 43, 44, 45, 111, 112, 119, and 123. Those claims are reproduced in Schedule A to these reasons. As I have noted above, Apotex challenged the validity of these claims on a number of grounds.

[14]

The development of the claimed invention 73 My findings concerning the development of the claimed invention are as follows. 74 Otsuka first synthesised aripiprazole in the laboratory in 1987. Within Otsuka, aripiprazole was designated as OPC-14597. The method by which aripiprazole was synthesised in the laboratory is disclosed in the 141 application. 75 From 1993, Otsuka made aripiprazole on a manufacturing scale by the following process (referred to in the evidence as Route I): crude aripiprazole was recrystallised from ethanol solution constituting 80% ethanol and 20% water to create aripiprazole hydrate; aripiprazole hydrate was dried at 80°C to make anhydrous aripiprazole; and the anhydrous aripiprazole was milled. 76 Otsuka began its Phase I clinical study on aripiprazole in Japan in 1990 and in the United States of America in 1993. 77 Aripiprazole manufactured industrially by Route I was found to be hygroscopic and this hygroscopicity posed a significant dissolution problem. Otsuka discovered in the course of conducting in vitro testing of aripiprazole tablets that there was a decrease in the dissolution rate of the aripiprazole. The decrease was caused when the aripiprazole, which initially existed as an anhydrate when manufactured, gradually turned into a hydrate as it absorbed moisture. The aripiprazole tablets with a decreased dissolution rate had lower bioavailability in patients with low levels of gastric acid compared to tablets with no decreased dissolution. 78 Otsuka commenced a research project to endeavour to solve this problem. 79 At the time that the Aoki article was submitted for publication (30 July 1996), Mr Aoki was continuing to investigate possible solutions to the hygroscopicity problem. In July 1996, he focused on a specific step in the manufacturing process of anhydrous aripiprazole. The step was to redry the aripiprazole after milling. Mr Aoki decided to investigate the effect of varying the conditions employed in drying the milled anhydrous aripiprazole. He ran experiments using the following temperatures and times: 80°C for 4, 8, 24 or 48 hours; 100°C for 4, 8, 24 or 48 hours; and 120°C for 4, 8, 24 or 48 hours. 80 He tested the results using a rigorous hygroscopicity test in which the samples were exposed to conditions of 40°C and 100% relative humidity for two weeks (the two-week test). 81 Mr Aoki's evidence, which I accept, was that he did not know in advance whether any of the conditions would successfully produce aripiprazole with low hygroscopicity. 82 In September 1996, he found that: milled anhydrous aripiprazole redried at 80°C for 4, 8, 24 or 48 hours all showed hygroscopicity under the two-week test; milled anhydrous aripiprazole redried at 100°C and 120°C for 4, 8, 24 or 48 hours showed no hygroscopicity under the two-week test; redrying at a temperature of at least 100°C was sufficient to resolve the dissolution problem; and the two-week test was reliable, based on the consistent hygroscopicity and dissolution results which were produced when the same sample was submitted to this test. 83 These discoveries were later applied to a modified method of preparing anhydrous aripiprazole industrially. This modified method (referred to in the evidence as Route II) comprised the following steps: crude aripiprazole was recrystallised from ethanol solution constituting 80% ethanol and 20% water to create aripiprazole hydrate; the hydrate was milled; and the milled product was dried at 100°C for 24 hours. 84 In July 1999, Otsuka conducted hygroscopicity testing of aripiprazole at conditions of 60°C at 100% relative humidity for 24 hours. This is the test referred to in claim 12 of the patent. It provided an alternative and equivalent hygroscopicity test to the two-week test. The advantage was that this test was more expedited so Otsuka could check quickly, during the manufacturing process, whether the process was going stably and whether the product had any problems with hygroscopicity. The test provided a distinct advantage in that it enabled Otsuka to ascertain within 24 hours whether the aripiprazole it had manufactured would have low hygroscopicity under the two-week test. 85 At the time that this test was devised, Otsuka was using Route II to manufacture aripiprazole. Otsuka found, applying this new test, that the Route II process produced aripiprazole that had low hygroscopicity. 86 Mr Aoki gave evidence that, at or around the time of publication of the Aoki article and the Aoki poster, he had not tested the hygroscopicity of the type 1 crystals referred to in the article and the poster using the new test. However, shortly before Japanese Patent Application No. 2001-290645 was filed (on 25 September 2001), he conducted such testing on those crystals. The method he used to prepare those crystals and the results he obtained on testing them are reported as Reference Example 2 in the complete specification. The results show that these crystals did not exhibit low hygroscopicity. 87 From September 1996 onwards, Otsuka performed additional work on how to manufacture aripiprazole in the form of Crystals B consistently on an industrial scale. These investigations were completed prior to finalising the filing of the Japanese equivalent of the patent. This work included the following: In 1997, Otsuka made its first manufacturing run of Crystals B on an industrial scale. In January 1998, it made a second manufacturing run of Crystals B on an industrial scale. In 1998, during a manufacturing run of Crystals B on an industrial scale, Otsuka discovered the presence of Type III Crystals (referred to in the complete specification as Crystals C) within the manufactured product. In 1999, it was confirmed that the presence of Crystals C was caused by over-grinding. In 1999, during the manufacturing of Crystals B on an industrial scale, Otsuka discovered the presence of Type IV Crystals (referred to in the complete specification as Crystals D) within the manufactured product. In July 2000, Otsuka found that this could be controlled by additional drying at 120°C for two hours. 88 A number of Otsuka's confidential internal documents relating to its investigations were tendered and Mr Aoki was cross-examined on these documents. It is not necessary to set out the detail of this evidence. The internal documents reveal that Otsuka faced a number of other challenges in relation to the production of aripiprazole and carried out other investigations. These facts are not directly relevant, however, to the issues raised in this proceeding.

[15]

Polymorphism and hygroscopicity 89 An important aspect of Apotex's case is its contention that hygroscopicity is an inherent property of a given crystalline form of a given compound. Apotex contended that a given crystalline form of aripiprazole will have, inevitably, a given hygroscopicity so that the disclosure of a particular crystalline form of aripiprazole will also inevitably disclose certain of the physicochemical properties of that form, including its hygroscopicity. More specifically, Apotex contended that, if anhydrous aripiprazole crystals have an XRPD spectrum that is substantially the same as the XRPD spectrum shown in Figure 5 of the complete specification, those crystals will inevitably have low hygroscopicity, namely, a moisture content of 0.40% or less after being placed for 24 hours in a closed container maintained at a temperature of 60°C and a humidity level of 100%. 90 In this connection, Apotex relied on, among other things, a finding made by me when dealing with Otsuka's earlier application to amend the patent. In the amendment reasons at [17], I said: [17] For present purposes I am satisfied on the evidence that a person skilled in organic chemistry in 2002 would have understood, on reading the specification as filed, that properties (7)-(11) are the essential physicochemical properties that define the crystalline form of Crystals B. 91 The properties designated as (7) to (11) in that quotation are the characteristics identified in the same way in the passage from the complete specification quoted in [54] above. That finding was based on the evidence then before me which was, relevantly, Professor Easton's first affidavit, which was also read at the present hearing. I did not have, at that time, the benefit of the far greater body of evidence now before me (including further evidence from Professor Easton), which I will now discuss. Before doing so, I should state that, had I had the benefit of the present evidence when dealing with the amendment application, I would have expressed myself differently when making my finding about what the person skilled in the art would have understood to be the physicochemical characteristics of Crystals B when reading the complete specification. That would not, however, have led me to reach any different conclusion on the fate of the amendment application. 92 Dr Rowe, who was called by Apotex, gave evidence in his first affidavit that, if the XRPD analysis indicates that a compound is in a particular crystalline form, then the compound will have the physicochemical properties of that particular crystalline form. He said that, if a sample of aripiprazole is determined by XRPD to be Crystals B, then it would have certain physicochemical properties, namely, the characteristics by which Crystals B are defined in claim 12 of the patent, including the low hygroscopicity characteristic to which I have referred. It appears that Apotex's contention that hygroscopicity is an inherent property of a given crystalline form of a given compound is derived from Dr Rowe's evidence. 93 Dr Rowe later qualified this evidence in light of certain observations made by Professor Withers in an affidavit that was read in the applicants' case. 94 Professor Withers gave evidence that the hygroscopicity of a sample of crystals cannot be determined by the XRPD pattern of those crystals alone. He said: 83. [The applicants' solicitors] asked me whether I am able to comment on the hygroscopicity of a sample based on its XRPD pattern alone. The answer is no. The hygroscopicity of a crystal needs to be determined experimentally. It can arise from (i) the absorption of water from the surrounding atmosphere into its crystal structure in a regularly repeating manner to form a hydrate, and/or (ii) the adsorption of water onto the surface of the crystal. In the former case, the measured hygroscopicity will be relatively high and the XRPD pattern of the hydrate will differ from that of the initial sample. In the latter case, the measured hygroscopicity will be relatively low and the XRPD patterns before and after adsorption will not differ. XRPD provides a fingerprint of the internal crystalline form, but not of its surface properties. 84. It is possible for two crystal samples to have the same XRPD patterns but different surface properties such that the hygroscopicity of those samples is different. Hygroscopicity is determined by, amongst other things, the surface properties, for example the particle size, of a crystal. 85. If I were to compare the XRPD patterns of two samples and consider that they were the same, or substantially the same, I would conclude that the two samples were of the same crystalline form. I would not be able to conclude that those two samples necessarily have the same hygroscopicity. 95 As I have noted, Professor Withers was not cross-examined. 96 When Dr Rowe responded to these statements in his second affidavit, he accepted that XRPD data alone does not confirm the hygroscopicity of a sample. He said, however, that, if the hygroscopicity of a crystalline form identified by XRPD analysis is known, the "inherent hygroscopicity" of the sample will be the same. This statement was itself qualified by Dr Rowe's acceptance that particle size can affect the amount of water a sample adsorbs, because the surface area of a particle can vary with particle size. In this connection, Dr Rowe said: 10. Professor Withers says … that the hygroscopicity of a crystal sample is determined by surface properties including the particle size, of a crystal. I agree that particle size can affect the amount of water a sample adsorbs because the surface area of a particle can vary with particle size … However, the water content of crystalline samples of the same compound, in the same crystalline form and with the same surface area will be the same (assuming they are exposed to the same environmental conditions), as hygroscopicity is an inherent property of the crystalline form. Smaller particles generally have greater surface area than larger particles, and this can mean that more water will be adsorbed … to the surface of the smaller particle than a larger particle. In my experience, the variation in adsorbed water content resulting from different particle sizes is usually very minor in samples of pharmaceutical material (particularly as samples are normally comprised of particles with a range of different surface areas). Surface area may also affect the rate of conversion of a hygroscopic compound that is susceptible to conversion to a hydrate, or is deliquescent … 97 In his first affidavit, Professor Easton said that XRPD, TGA/DTA, DSC, adsorption bands on the IR (KBr) spectrum, and hygroscopicity measurements afford information that is "characteristic of a particular crystal form". When referring to the claims of the patent before amendment, Professor Easton said: 4.13 Upon reading the [patent] and its claims, I understood the claims relating in particular to Anhydrous Aripiprazole Crystals B to cover the crystal form with the Characteristics set out on pages 23 to 24 of the [patent] and, to the extent that any of the claims of the [patent] repeated one of the Characteristics, I considered that repetition to be redundant. That is, I did not see any differences in what is claimed between, for example, claim 12 of the Amended Claims and unamended claims 14, 16, 17 and 18 of the [patent]. Further, to the extent that the claims of the [patent] introduced any additional characteristics not covered by the Characteristics set out on pages 23 to 24, I considered the relevant claims to be narrower in scope and to cover a crystal form with more narrowly defined characteristics than the Characteristics set out on pages 23 to 24 of the [patent]. For example, claim 13 of the [patent] specifies a further reduced moisture content of 0.10% or less, which is narrower than the 0.4% figure set out on page 24 of the [patent]. 98 In his second affidavit, when referring to the properties of polymorphs of the same chemical compound, Professor Easton also accepted that the form of the material affects various of its properties, including its hygroscopicity. He said, however, that hygroscopicity is also determined by characteristics such as crystal size and surface area. 99 In cross-examination, he also suggested that, when a crystal structure takes up water, there may be a relationship between adsorption and absorption: when a material converts from an anhydrous form to a hydrate, it generally goes through an adsorption phase first, which will be affected by the smoothness of the crystal surface, including whether there are fractures in it. He said that there has to be a localised restructuring to convert the anhydrous form to the hydrous form, "which basically involves one [disassembling] and then another one re-assembling". 100 Professor Prestidge also gave evidence on this subject. When asked in cross-examination whether the degree of hygroscopicity of a polymorph under given constant conditions is a property of that polymorph, Professor Prestidge said that the moisture uptake of the polymorph can be a "surface phenomenon" (that is, the adsorption of moisture on the surface of the crystals) or a "bulk phenomenon" (that is, the absorption of moisture into the bulk of the crystals). He also said that there may be factors other than polymorphic form that can influence the moisture uptake of a crystal. He said that, if there are changes in particle size and surface area in two samples of the same polymorph, the samples may have different levels of moisture uptake. Professor Prestidge agreed, however, that two samples of the same polymorph with the same surface properties and the same bulk properties should have the same moisture uptake. 101 Importantly, however, he gave this evidence: … [P]olymorphic form is one of the factors that determines hygroscopicity, would you agree?---One of the factors, yes. Yes. And if I keep surface area constant in the way you've just described it, a given polymorph will have a given hygroscopicity?---There are subtleties in this field, because you have a bulk crystal and the surface chemistry can be subtly different. So the techniques we use here to characterise the polymorphic form, as in the infrared spectroscopy and the XRPD, they measure bulk properties. Moisture uptake can be controlled by a surface phenomenon which is a little bit more subtle in some ways … 102 He explained that there may be a number of physical phenomena that can influence the capacity of a substance to adsorb moisture. When asked about how the surface properties of a polymorph might be affected by the way in which the polymorph is produced, Professor Prestidge said: … [T]he way in which a polymorph is formed controls the bulk properties as measured by the various crystal structures, but other phenomenon, for instance, grinding or other physical thermal properties, may change - the molecules that are at the absolute surface of a crystal are different from those in the bulk. So there is - there are a number of physical properties - physical phenomenon that may influence that. 103 He also gave this evidence: If an anhydrous form converts to a hydrated [form], and that's the process where we get absorption of water, that's likely to be quite a substantial moisture uptake. In terms of percentages, I think we're talking a number of percentages, so it's - so for instance, we go from a monohydrate, each molecule of the drug will absorb one molecule of water, so you can do a calculation there of what the [likely] uptake is. With respect to adsorption, the physical effect, that depends on the particle size. If you've got materials of this sort of - this sort of particle size, in the 10s of microns, it's likely to be lower values. So the amount of moisture uptake on the adsorption onto these sized particles is likely to be a fraction of a per cent. 104 When dealing with the surface properties of a particle other than its surface area, Professor Prestidge said: And what other surface properties do you have in mind other than surface area?---It's around the molecule - if the molecule is orientated slightly differently at different crystal faces, that can play a role. You can have the same polymorph - so, yes, literally, it's around subtle molecular structural changes that can occur at the surfaces of crystals which may play a role. 105 Because of a lack of data, Professor Prestidge was unable to express a view concerning the percentage of water uptake that might be caused by such changes. 106 The evidence of these witnesses shows that the hygroscopicity of a crystalline compound is not just a function of its crystalline form, but may be influenced by particle size and, in subtle ways, by the surface properties of the polymorph in question. Professor Prestidge's evidence suggests that, as a general statement, water uptake by adsorption is likely to be significantly less than water uptake by absorption, when the particle size is in the tens of microns, such as the crystals claimed in claim 16. Even so, I do not think that the possible influence of these factors can simply be ignored. 107 In submissions, Apotex placed considerable reliance on Otsuka's internal memoranda and experimental reports to advance explanations of why conventional anhydrous aripiprazole crystals, such as type 1 crystals and type 2 crystals, exhibit the hygroscopicity problem described in the complete specification. Apotex submitted that this material shows that the cause of the problem was not an inherent property of type 1 crystals that is not also possessed by Crystals B. 108 My own reading and analysis of these documents, assisted by Mr Aoki's cross-examination on them, does not persuade me that, at the time the patent was applied for, Otsuka itself had any definite understanding of why conventional anhydrous aripiprazole crystals were "significantly hygroscopic". It gave consideration to multiple possible causes for the problem. Its consideration was based, in part, on the empirical work it had undertaken and was continuing to undertake to gain a greater understanding of the physicochemical properties of crystalline aripiprazole. Its research, carried out over an extended period of time, ultimately led it to conclude that, by undertaking certain processing steps, it was possible to achieve crystalline aripiprazole that exhibited low hygroscopicity in terms of the moisture content described in the complete specification as being one of the essential characteristics of Crystals B. 109 For reasons that I will later explain, I am satisfied on the balance of probabilities that Crystals B and type 1 crystals have the same crystalline form. However, if that is the case, Otsuka's experimental work shows that the hygroscopicity of a given crystalline form of aripiprazole, under certain storage criteria, will vary depending on whether the aripiprazole in that crystalline form is subjected to certain processing steps, such as those with which Mr Aoki was involved, including heating at certain temperatures. The complete specification reveals that type 1 crystals do not have the same hygroscopicity as Crystals B under certain conditions unless the type 1 crystals undergo heating at specific temperatures. I do not accept, therefore, that a given crystalline form of aripiprazole will have, inevitably, a given hygroscopicity. It follows that I do not accept that the disclosure of a particular crystalline form of anhydrous aripiprazole will also inevitably disclose its hygroscopicity in that form. 110 In this connection, Crystals B will maintain a water content of 0.40% or less after 24 hours in a closed container set at a temperature of 60°C and a humidity level of 100%. Type 1 crystals will exhibit a higher water content under those conditions if they do not undergo suitable heating. 111 The complete specification illustrates suitable heating through, for example, the comparison between the Reference Examples and the Examples. The complete specification makes clear that, in this connection, "heating time and heating temperature are inversely related". It teaches that Crystals B "can be prepared with certainty" by heating Hydrate A for about 18 hours at 100°C and then heating it for about 3 hours at 120°C. Thus, although the complete specification proposes a process of preparation of crystalline aripiprazole that involves milling conventional hydrate to produce Hydrate A as an intermediate, a particular heating step is still required as part of the process for obtaining Crystals B. 112 It is clear that, although a process using Hydrate A is stated to be the most efficient, a process using Hydrate A as an intermediate is not the only process by which Crystals B can be obtained. The complete specification teaches that crystalline aripiprazole having the characteristics of Crystals B can be obtained by heating conventional anhydrous aripiprazole crystals at 90°C to 125°C over a period of time that is generally about 3 to 50 hours: see, for example, the penultimate paragraph in the passage of the complete specification quoted in [55] above. I stress this point because, time and again, Apotex's submissions focused on what it said was Otsuka's invention of preventing the conversion of type 1 crystals to monohydrate by milling hydrous crystals to produce Hydrate A and then drying them. These submissions, which present an incomplete picture of what is disclosed in the complete specification, if not treated with care, serve to distract attention from the full description of the invention and the processes by which Crystals B can be obtained. 113 This is illustrated by comparing the results of Reference Example 1 with the results of Example 5. Reference Example 1 shows the preparation of conventional anhydrous aripiprazole crystals. The process includes drying raw aripiprazole for 40 hours at 80°C to obtain anhydrous aripiprazole crystals. It is known from the complete specification that these are type 1 crystals. When these crystals were left for 24 hours in a dessicator set at a temperature of 60°C and a humidity level of 100%, they exhibited hygroscopicity of 3.28%. When, in Example 5, the same crystals were subjected to additional heating for 50 hours at 100°C and then subjected to the same temperature and humidity testing, they exhibited hygroscopicity of 0.04%. Given the differences in hygroscopicity revealed in this comparison, the product of Example 5, but not the product of Reference Example 1, had the characteristic low hygroscopicity of Crystals B. The results of Reference Example 1 can also be compared with the results of Example 6, where the crystals from Reference Example 1 were heated for 3 hours at 120°C. When subjected to the same temperature and humidity testing, these crystals exhibited hygroscopicity of 0.04% (the same as for the product of Example 5). The crystals from Example 6 therefore had the characteristic low hygroscopicity of Crystals B. Reference Example 1 and Examples 5 and 6 do not involve the preparation of anhydrous aripiprazole crystals using Hydrate A as an intermediate, but they do relate to anhydrous aripiprazole crystals having the same crystalline form. 114 Apotex challenged the role that adsorption might play in explaining the hygroscopicity of conventional anhydrous aripiprazole crystals. It submitted that "the recently articulated surface properties/adsorption case theory of [the applicants] has absolutely no basis in the [p]atent or in Otsuka's documents". I do not accept that that statement is entirely correct when viewed from the perspective of Otsuka's internal documents. More relevantly, Apotex submitted that the complete specification "makes no mention whatsoever of surface properties or adsorbed water". 115 The validity of the patent does not turn on the applicants establishing the scientific reason why the invention works. Apotex did not plead that the results of the experimental work recorded in the complete specification are false. If, as Apotex contended, and as I accept, Crystals B and type 1 crystals have the same crystalline form, then the results of these experiments show that by suitable heating, as exemplified in the complete specification, something happens that causes the crystals to have the low moisture content that is characteristic of Crystals B as claimed in claim 12 of the patent. This conclusion is supported by the work disclosed in Otsuka's internal memoranda and experimental reports to which I have referred. What that "something" is, is neither here nor there. An inventor does not need to know why his or her invention is afforded utility in order for that invention to be patentable. Moreover, it does not matter that the invention can be achieved by various, indeed by many, means. All that a patentee in Otsuka's position is required to do is to describe the invention fully, including by giving the best method known to the patentee of performing the invention. Apotex has not challenged the validity of the patent on the basis that Otsuka has failed in that requirement. 116 It is necessary to deal with a further aspect of Apotex's submissions. Apotex sought to rely on what it saw as a distinction between the properties of a given crystalline compound, specifically, aripiprazole as Crystals B, and the properties of samples of that compound. Its purpose in doing so appeared to be to focus attention on what the complete specification discloses as the properties of Crystals B, not to what the properties of Crystals B might be found to be when various manufactured samples are compared. The stated end point of this reasoning was that the complete specification refers to and claims a particular crystalline form of aripiprazole - Crystals B. 117 I do not accept that reasoning. In my view, the distinction which Apotex sought to draw is not meaningful in the present context. The complete specification is a practical document giving practical information about the manufacture and use of aripiprazole as a pharmaceutical product. It necessarily refers to the problem of hygroscopicity associated with conventional anhydrous aripiprazole crystals that have been prepared on an industrial scale for pharmaceutical processing and formulation. The complete specification's description of the invention by reference to examples necessarily involves a consideration of, and comparison between, samples of anhydrous aripiprazole crystals. 118 More importantly, I do not accept that the complete specification merely refers to and claims a particular crystalline form, or crystalline forms, of aripiprazole. It describes and claims crystalline aripiprazole having certain characteristics. When the complete specification refers to the form of aripiprazole crystals, it is not referring only to their crystalline form but more generally to their form as crystals defined by cumulative characteristics. In the case of Crystals B, one of these characteristics is that the crystals have an XRPD spectrum that is substantially the same as the spectrum disclosed in Figure 5; that is, they have a particular crystalline form. Another characteristic is that, in that crystalline form, the crystals must have low hygroscopicity in a specific sense: their moisture content is 0.40% or less after being placed for 24 hours in a closed container maintained at a temperature of 60°C and a humidity level of 100%. 119 This returns me to the finding that I made at [17] of the amendment reasons. With the benefit of all the evidence now before me, that finding would have been better expressed by saying that, at the relevant time, the person skilled in the art would have understood, on reading the complete specification as filed, that properties (7) to (11) are the essential physicochemical properties that define Crystals B.

[17]

The 141 application 121 The 141 application was made publicly available on 9 May 1990. It states that the invention relates to novel carbostyril derivatives and their salts. It discloses, among other things, aripiprazole and its utility in treating schizophrenia. In that connection, it also discloses processes for the preparation of aripiprazole and pharmaceutical compositions containing aripiprazole. 122 Example 1 in the 141 application is of particular significance in the present case. The parties accept that this example is relevantly the same as Example 1 in the Japanese unexamined patent publication referred to in the complete specification of the patent. Lines 5 to 15 of Example 1 are as follows: A suspension of 47 g of 7-(4-bromobutoxy)-3,4-dihydrocarbostyril, 35 g of sodium iodide with 600 ml of acetonitrile was refluxed for 30 minutes. To this suspension was added 40 g of 1-(2,3-dichlorophenyl)-piperazine and 33 ml of triethylamine and the whole mixture was further refluxed for 3 hours. After the solvent was removed by evaporation, the residue thus obtained was dissolved in chloroform, washed with water then dried with anhydrous magnesium sulfate. The solvent was removed by evaporation, and the residue thus [obtained] was recrystallized from ethanol twice, to yield 57.1 g of 7-{4-[4-(2,3-dichlorophenyl)-1-piperazinyl]butoxy}-3,4-dihydrocarbostyril. Colorless flake crystals Melting point: 139.0 -139.5° C. 123 Example 1 continues with further directions for the preparation of four salt forms of aripiprazole: a hydrochloride, a sulphate, a fumarate, and a maleate. However, Apotex's case of anticipatory disclosure rests on that part of Example 1 (lines 5 to 15) quoted above. 124 The 141 application discloses other examples of prepared compounds (Examples 2 to 14). These compounds are not aripiprazole, but other carbostyril derivatives. 125 The parties relied on other disclosures in the 141 application which, they argued, would have provided contextual information to assist the person skilled in the art to understand, before 25 September 2001, how to carry out the instructions contained in lines 5 to 15 of Example 1. The significance of these disclosures lies in whether, and to what extent, they would have influenced the person skilled in the art, beyond the way in which lines 5 to 15 are expressed, to attempt to remove ethanol if found in the crystals after carrying out those particular steps. As will later become clear, this is an important consideration in arriving at a conclusion on whether lines 5 to 15 of Example 1 are an anticipatory disclosure of the invention claimed in claim 12 of the patent. 126 The relevant disclosures are as follows: The further instructions in Example 1 show that ethanol was used with the crystals obtained from the process in lines 5 to 15 to prepare each of the four salt forms of aripiprazole. Pharmaceutical testing was carried out on mice using certain of the compounds produced from carrying out the examples. The test compound deriving from Example 1 was administered in its free base (that is, non-salt) form. This indicates that the aripiprazole was not a solvate (specifically, not an ethanolate). In other words, any ethanol remaining the crystals after synthesis was removed. The 141 application does not specify the form in which the test compound obtained from following the directions in lines 5 to 15 of Example 1 was administered. However, Professor Black accepted that a solution would have been used. The 141 application discloses that ethanol is one of the substances that can be used as a binder or as a carrier in various pharmaceutical preparations containing aripiprazole, including in solutions. A pharmaceutical composition for a tablet formulation is disclosed containing aripiprazole (5 mg) in its free base form, combined with starch (132 mg), magnesium stearate (18 mg), and lactose (45 mg). The fact that 5 mg of aripiprazole is specified indicates that aripiprazole is not present in the exemplified formulation as a solvate (specifically, as an ethanolate).

[18]

The Aoki article 127 The Aoki article was made publicly available at the Symposium on or shortly after 6 October 1996. It reports on a study that was conducted to examine the effect of grinding on the physicochemical properties of crystalline aripiprazole. The investigation was carried out using XRPD, water content measurement, and DSC. 128 The Aoki article teaches that the crystal size distribution of a drug substance may affect its bioavailability through differing dissolution rates, and that crystal size distribution is usually adjusted by grinding. However, grinding can destroy the crystalline structure of a substance, resulting in a change to its bioavailability. Therefore, it is important to examine and understand the effect of grinding on the physicochemical properties of crystalline drug substances. 129 The Aoki article discloses three different crystalline forms of aripiprazole, which it designates as type 1, type 2, and type 3, respectively. Type 1 crystals are disclosed as anhydrous aripiprazole that has been recrystallised from ethanol solution. The melting point of these crystals is said to be 140°C. The Aoki article discloses that type 1 crystals are converted to type 2 crystals by heating type 1 crystals at 130°C to 140°C for 15 hours. The melting point of the type 2 crystals is said to be 150°C. When these polymorphs are recrystallised from alcoholic solvent in a solution containing water up to 20°C (v/v), the crystals are converted to hydrous crystals (type 3 crystals). The type 3 crystals are converted to type 1 crystals by heating them at 80°C. When heated at this temperature, the resulting type 1 crystals are stable and do not convert to type 2 crystals. 130 The Aoki article includes, as Figure 3, the XRPD patterns of each crystal type, although the depicted patterns are not shown by reference to measurements. Only the shape of each pattern is given. 131 The Aoki article also includes, as Figure 4a, the XRPD pattern of type 1 crystals during grinding and, as Figure 4b, the effect of grinding time on the intensity of the endothermic peak exhibited by type 1 crystals at 150°C as measured by DSC. 132 The Aoki article discloses that, with an increase in grinding time, the number of peaks in the XRPD pattern of the type 1 crystals does not change. However, the intensity of the peaks decreases as the grinding time increases. Moreover, the intensity of the endothermic peak of type 1 crystals increases as the grinding time increases. 133 Under the heading "Discussion", the Aoki article says that aripiprazole has two types of anhydrous polymorphs. It opines that the changes in melting point observed by grinding aripiprazole might be caused by the presence of these two polymorphs. However, it says that XRPD analysis "denied" crystal transformation during grinding. The Aoki article continues: But, it is also true that something occurs during grinding, because the intensity of powder X-ray diffraction decreased during grinding and the intensity of the endothermic peak at 150°C in DSC increased. 134 The Aoki article then offers the following explanation: Type 1 is probably converted to noncrystalline state by grinding. The noncrystalline product produced could be easily transformed to Type 2. The possible mechanism of crystal polymorph transformation through grinding is present in Fig.5. 135 Figure 5 is a diagram showing that the type 1 anhydrous crystals are partially converted to a non-crystalline form by grinding. When heated, the non-crystalline forms are transformed to type 2 anhydrous crystals. 136 The Aoki article concludes by stating, with reference to Figure 5, that the authors have "proposed a model of crystal polymorphs transformation via noncrystalline as an intermediate …".

[21]

Introduction 143 In The General Tire & Rubber Company v. The Firestone Tyre and Rubber Company Limited and Others [1972] RPC 457, the Court of Appeal, when discussing the circumstances in which a claimed invention will be anticipated by a prior published document, said (at 485-486): … If the prior inventor's publication contains a clear description of, or clear instructions to do or make, something that would infringe the patentee's claim if carried out after the grant of the patentee's patent, the patentee's claim will have been shown to lack the necessary novelty, that is to say, it will have been anticipated. The prior inventor, however, and the patentee may have approached the same device from different starting points and may for this reason, or it may be for other reasons, have so described their devices that it cannot be immediately discerned from a reading of the language which they have respectively used that they have discovered in truth the same device; but if carrying out the directions contained in the prior inventor's publication will inevitably result in something being made or done which, if the patentee's patent were valid, would constitute an infringement of the patentee's claim, this circumstance demonstrates that the patentee's claim has in fact been anticipated. If, on the other hand, the prior publication contains a direction which is capable of being carried out in a manner which would infringe the patentee's claim, but would be at least as likely to be carried out in a way which would not do so, the patentee's claim will not have been anticipated, although it may fail on the ground of obviousness. To anticipate the patentee's claim the prior publication must contain clear and unmistakeable directions to do what the patentee claims to have invented: Flour Oxidizing Co. Ltd. v. Carr & Co. Ltd. ((1908) 25 R.P.C. 428 at 457, line 34, approved in B.T.H. Co. Ltd. v. Metropolitan Vickers Electrical Co. Ltd. (1928) 45 R.P.C. 1 at 24, line 1). A signpost, however clear, upon the road to the patentee's invention will not suffice. The prior inventor must be clearly shown to have planted his flag at the precise destination before the patentee. 144 In the present case, Apotex contended that lines 5 to 15 of Example 1 of the 141 application contain clear directions to do or make something that would infringe claim 12 of the patent, if carried out after the grant of the patent. It sought to make good this contention by carrying out an experiment to reproduce that part of Example 1. The applicants responded by carrying out their own experiment. This aspect of the case involved a considerable amount of evidence, which included significant criticism of how each experiment was designed to be carried out and how each was in fact carried out. By the commencement of the hearing, the area of dispute requiring resolution had become significantly more focused. In these circumstances, I will confine my summary of the evidence, and my conclusions on it, to reflect only those matters that featured in the case finally presented by the parties as to whether the directions in lines 5 to 15 of Example 1 of the 141 application anticipate claim 12 of the patent. If claim 12 is not anticipated by the relevant directions, it is accepted that the other challenged claims are not anticipated.

[22]

The McGeary protocol 145 The solicitors for Apotex approached Associate Professor McGeary to prepare a written protocol detailing how, practically, he would have carried out the directions in lines 5 to 15 of Example 1 in 2001 (the McGeary protocol). 146 In his first affidavit, Associate Professor McGeary said: 76 Example 1 describes a method to synthesise aripiprazole. Example 1 provides an abbreviated description of this method, but in my opinion, there is nothing unusual in the degree of abbreviation of Example 1. I consider that Example 1 provides enough information for a person experienced in organic synthesis to follow the procedure and synthesise aripiprazole. 77 Nothing in Example 1 strikes me as being particularly unusual or deviating from standard techniques. In my opinion it describes a straightforward process, and I would have considered this to be a straightforward process prior to September 2001. With organic syntheses, many variables are considered when developing a method of synthesising a desired product, for example, the reaction solvent, the duration of the reaction steps, and temperatures. I take it at face value that the conditions described in Example 1 work. In my opinion there is nothing unusual regarding the reagents or solvent used in the reaction. 78 Minimal characterisation of aripiprazole is provided in the [141 application] compared with something that would appear in, for example, a peer reviewed journal that reports organic syntheses. In my experience, reputable peer reviewed journals dealing with organic syntheses typically require melting point, 1H-NMR and 13C-NMR spectra, and either mass spectral or elemental analysis data to be provided to confirm the identity of a compound if it has not previously been reported. If the compound has previously been reported, less information may be provided, for example, I would expect the melting point to be reported with a cross-reference to a publication of the compound that provides more detailed analytical information. The only characterisation given in the [141 application] is the description of the product as 'colorless flake crystals' (which indicates that no (or minimal) coloured impurities are present in the aripiprazole) with a melting point of 139.0-139.5°C. This relatively narrow melting point range suggests to me that the method reported in Example 1 results in the production of substantially pure aripiprazole … 147 Associate Professor McGeary referred to the fact that, in the 141 application, Example 1 proceeds with further instructions to produce certain salt forms of the compound. He said, however, that nothing in those instructions had an impact on what is described in lines 5 to 15 of Example 1 or on how he would have carried out those instructions in 2001. Nevertheless, he noted that the further instructions provided information about the solubility of aripiprazole in ethanol, namely, that 1 g of aripiprazole was dissolved in 20 ml of ethanol. 148 The McGeary protocol is lengthy. It is not necessary for me to set it out in its entirety because the focus of the debate between the parties came to be centred on the steps proposed by Associate Professor McGeary following the second recrystallisation with ethanol referred to in lines 5 to 15 of Example 1. Steps 4.7 and 4.8 of the McGeary protocol gave procedures for carrying out the first and second recrystallisations, respectively. The McGeary protocol then proposed the following steps as steps 4.9 and 4.10, which are of central significance to this part of Apotex's case: 4.9 If NMR analysis indicates that traces of solvent remain further drying steps will be undertaken to remove the residual solvent. Further drying will be undertaken in a vacuum desiccator at room temperature. Samples will be taken at appropriate intervals and analysed by NMR for the presence of solvent. The solvent can be so firmly held that it cannot be completely removed in a vacuum desiccator at room temperature. If NMR indicates that traces of solvent remain, I would conclude that some ethanol is bound within the crystals (as opposed to being retained on the surface of the crystals). In this circumstance further drying will be undertaken in an oven at between 80°C and 100°C for an initial overnight period (or equivalent). Samples will be taken at appropriate intervals and analysed by NMR for the presence of solvent. 4.10 Once dry, the yield, physical appearance and melting point of the product are determined. The product will be characterised by the methods described in Section 6. The balance of product not used will be retained. [Footnotes omitted] 149 Section 6 of the McGeary protocol set out how the final product of the McGeary protocol would be analysed to ascertain its characteristics. 150 In his first affidavit, Associate Professor McGeary spoke of the need to isolate and purify the reaction product. He gave several reasons for this, including that aripiprazole is used as a pharmaceutical and that impurities in the pharmaceutical product may have unwanted, including toxic, side effects and adversely affect the bioactivity of the compound. He also explained the need to isolate the pure compound for characterisation to provide control data that can be used for meaningful comparison with data from later samples. He noted, in this connection, that the "narrow melting point range" identified in Example 1 (139.0°C to 139.5°C) suggests that the product of Example 1 "should be pure". 151 Associate Professor McGeary commented on the direction in Example 1 to recrystallise the residue from ethanol twice. He said that recrystallisation is a purification technique. With respect to Example 1, he said: 144 Example 1 directs me to recrystallise the residue from ethanol twice. Put another way, Example 1 describes a single solvent recrystallisation that is repeated. This suggests to me that a single recrystallisation does not result in sufficiently pure product. It is well known to me and other organic chemists that repeating a recrystallisation increases the purity of the resulting crystalline compound. Example 1 does not detail the specific steps taken to recrystallise the residue, however in my opinion there is nothing unusual in the degree of abbreviation used by the authors of the [141 application] in this regard. … 152 With respect to the use of ethanol to perform the recrystallisation, he said: 157 Example 1 specifies that ethanol is used as the recrystallisation solvent. Ethanol is commonly used as the recrystallisation solvent when purifying compounds for pharmaceutical use. Ethanol has a relatively low boiling point (approximately 80°C) which assists in its removal by evaporation. As long as the compound has a higher melting point, the compound is unlikely to be degraded by heating it to a suitable temperature to drive off any remaining ethanol. Ethanol is the intoxicating component of alcoholic drinks. As such, while it is bioactive, it is generally considered to be of low toxicity. 153 He noted that Example 1 does not specify the purity of the ethanol to be used. It was his opinion, however, that ethanol of at least 99% purity was "an appropriate choice". 154 Associate Professor McGeary explained why his protocol included steps 4.9 and 4.10: 161 Having performed the second recrystallisation and dried the recovered product for around 30 minutes under suction, the recovered product is tested by NMR to determine whether any ethanol remains as an impurity. … remaining ethanol is an impurity that must be removed from the aripiprazole to enable proper characterisation of the aripiprazole and for its use as a pharmaceutical. 162 If ethanol remains, it may have physically adhered to the crystals, ie, adsorbed on the surface of the crystal. I expect that drying the recovered product in a vacuum desiccator will remove ethanol that has adhered to the surface of the crystals. As described in paragraph 4.9 of the [McGeary protocol], I would firstly dry the recovered product in a vacuum desiccator if ethanol is detected by NMR analysis. This drying is to be monitored by NMR analysis. I would not expect this drying to remove any ethanol incorporated into the crystals as a solvate. 163 If NMR analysis indicates that drying in a vacuum desiccator has not removed all ethanol present. I would expect that any such ethanol could be removed by heating the recovered product at a temperature higher than the boiling point of the ethanol (roughly 80°C). As described in paragraph 4.9 of the [McGeary protocol], I would heat the product at between 80 and 100°C to remove this ethanol. It is difficult to predict how long this process will take, as in my experience, it could occur within a few hours, or several days. Paragraph 4.9 therefore provides for the dryness of the recovered product to be measured at appropriate intervals by NMR analysis. I use the term "dryness" in this context to refer to the recrystallised product being free of ethanol. 164 I would not be concerned if I didn't obtain a yield of 57.1g of product. In my experience, it is common for the highest yield obtained after running a particular method (at least) several times to be reported. It is also common in my experience for the yield obtained to be increased by performing a particular method several times, as an operator will get better at the method with repetition. 155 Example 1 of the 141 application refers to the product resulting from the steps in lines 5 to 15 as "colorless flake crystals". Associate Professor McGeary said: 165 In my experience, it is not unusual to have a slightly discoloured product. A tiny amount of an impurity can potentially change the colour of the product. I would only be concerned if a product that was expected to be colourless was brightly coloured. The structure of this compound suggests it should not be coloured and the observed appearance of the product of Example 1 states that the product is colourless. 166 I am not aware of the term 'flake crystal' being used by organic chemists to refer to crystals with any particular shape. In my opinion, this term may have been included as an incorrect translation of another term to English. It does however suggest that the crystals may be flat and thin as opposed to, for example, blocks or needles. I would primarily rely upon: (a) a comparison of the melting point reported in Example 1 and the measured melting point of the product, and (b) NMR data, to determine whether I obtain aripiprazole. 156 Associate Professor McGeary made the following general observation concerning his protocol: 168 The [McGeary protocol] accurately reflects how I would have undertaken the repetition of Example 1 of the [141 application] before September 2001. It reflects what I would have done in September 2001 as a matter of routine. In preparing the [McGeary protocol] I performed no tests. Nor have I validated the [McGeary protocol] by carrying out any aspect of it prior to undertaking the experimental repeat of Example 1. In my opinion the [McGeary protocol] does not contain any step that could reasonably be said to arise from my unique experience. Rather, the [McGeary protocol] reflects what, in my opinion, the typical organic chemist would do on being asked to undertake the repetition of Example 1 of the [141 application] in 2001.

[23]

Other evidence concerning the McGeary protocol 157 Professor Black said that lines 5 to 15 of Example 1 of the 141 application set out, at a high level, a process to produce aripiprazole. That part of Example 1 does not give specific reaction details. He said, however, that it contains sufficient details to allow an organic chemist with a standard knowledge of organic synthetic techniques "to proceed with confidence that they will be able to synthesise aripiprazole using this example". 158 He considered the reaction and purification processes described in lines 5 to 15 of Example 1 to be "very straightforward, and to involve techniques that are, and were before September 2001, routine". He observed that any compound to be used as a pharmaceutical, or in testing for suitability as a pharmaceutical, should be obtained in pure form. He expected, therefore, that the synthesis of the product according to lines 5 to 15 of Example 1 would yield an "analytically pure" compound, in the sense that standard analytical tests would not detect impurities in that compound. 159 Professor Black expressed the opinion that the McGeary protocol accurately represents how, using standard techniques, he and other organic chemists would have set out to repeat the process described in lines 5 to 15 of Example 1 of the 141 application in September 2001 in order to synthesise and purify aripiprazole. In his first affidavit, he said: 55. In my opinion, the techniques proposed to be used in the [McGeary protocol] were all routine techniques generally available to organic chemists in September 2001. Based on my experience in teaching organic chemistry and day to day interactions with organic chemists … I consider that these techniques would have been known by and available to organic chemists generally in September 2001. 160 With specific reference to section 4 of the McGeary protocol, he said: 70. In summary, based on my review of Section 4 of the [McGeary protocol] … it is my opinion, that the experimental steps set out in Section 4 of the [McGeary protocol] are an accurate way to reproduce Example 1 using techniques that were known to, and routinely used by, organic chemists in September 2001. In my opinion, based on my experience in teaching organic chemistry, carrying out research, and collaborating with organic chemists … I consider that other organic chemists would have also adopted this approach. 161 Professor Black referred to the instruction in Example 1 to perform two recrystallisations. He said that it was common in September 2001 to perform additional recrystallisations if undesired impurities remained in a product after recrystallisation. 162 With respect to step 4.7 in the McGeary protocol that provided for the first recrystallisation, Professor Black said that it was not necessary to confirm that all the solvent (in this case, ethanol) had been removed from the crystals because they were to be immediately redissolved in ethanol. He noted that step 4.8 of the McGeary protocol provided that the second recrystallisation was to be performed in the same manner as the first recrystallisation "refined as appropriate" by the experience of the first recrystallisation, and that the dryness of the product was to be confirmed by NMR spectral analysis. He said that, in his experience, it was routine in September 2001 to confirm that all solvent had been removed after the final recrystallisation, in order to obtain an analytically pure sample. 163 With respect to this step, Professor Black said: 90. If NMR analysis indicates ethanol remains, further drying of the crystals is required to remove that ethanol and recover the recrystallised product. I would expect to be able to remove that ethanol using the conventional techniques described in paragraph 4.9 of the [McGeary protocol]. Therefore, in September 2001, I would have routinely used the techniques described in paragraph 4.9 of the [McGeary protocol], ie, drying in a vacuum desiccator at room temperature, and if ethanol still remained, drying in an oven at between 80 and 100 °C (ie, above 78°C, the boiling point of ethanol). The removal of ethanol will be evident in routine NMR analysis of the crystals. In my opinion, based on my experience teaching organic chemistry, carrying out research in organic chemistry and collaborating with organic chemists … other organic chemists would have also adopted this approach. 164 In later paragraphs of these reasons, I will return to discuss some more detailed evidence given by Professor Black concerning appropriate drying steps that could be undertaken with respect to the product obtained from the second recrystallisation. 165 Professor Easton was of the view that there were several differences between the McGeary protocol and how he and other skilled organic chemists would have carried out the instructions in lines 5 to 15 of Example 1. 166 Significantly, in his second affidavit, Professor Easton said that there was nothing in lines 5 to 15 of Example 1, or elsewhere in the 141 application, that directed the person skilled in the art to carry out step 4.9 of the McGeary protocol. Professor Easton said: 21. I consider that there is nothing in Example 1 or the body of the [141 application] which directs a person to further dry the product, including in an oven between 80 °C and 100 °C for an initial overnight period (step 4.9 of the McGeary [p]rotocol), and I consider that that step would not have been taken by me, or other skilled chemists, if I were to try to repeat lines 5-15 of Example 1 for the reasons set out below. I see no basis for Associate Professor McGeary performing that step. 167 Professor Easton continued: 22. If I were writing a protocol for an experiment to reproduce the product of lines 5 - 15 of Example 1, I would not have included a step like step 4.9 of the McGeary [p]rotocol because it would not have been known to me, and I would have had no reason to assume, at the time of writing the protocol that the product after the second recrystallisation would have contained ethanol. (a) I would have expected that the suction in step 4.8 (which is a repeat of the corresponding process in step 4.7) would have removed any ethanol recrystallisation solvent. (b) Based on the method described in Example 1 and how I would have carried out that method, I would have had no reason to assume that the recrystallisation would produce an ethanol solvate (known as an ethanolate). 23. If NMR analysis were undertaken on the material, as is specified in step 4.8 of the McGeary [p]rotocol, and that analysis showed that ethanol was present in the material, I would have continued drying under suction or I would have air-dried the material to constant weight. 24. If subsequent NMR analysis were undertaken on the material, and that analysis still showed the presence of ethanol or water, it would have simply established that the product of lines 5 - 15 of Example 1 is a ethanolate or a hydrate. An ethanolate is the name for a solvate of which the solvent of the solvate is ethanol. A hydrate is the name for a solvate of which the solvent of the solvate is water. Obtaining an ethanolate at this stage is not an unexpected result; it is simply not known at the time of writing a protocol whether a non-solvate or a solvate will be obtained. 168 Professor Easton developed his views by the following reasoning. First, if the person skilled in the art continued to follow the rest of Example 1 to obtain the specified salt forms of aripiprazole, it would not matter that an ethanolate was obtained by following that part of Example 1 quoted above in [122]. He said that he would be less concerned with purifying and characterising "that intermediate product". The instructions of Example 1 for the preparation of the salt forms of aripiprazole include the step of dissolving the crystals in ethanol. 169 Secondly, in his experience, the chemical names for the ethanolate and the non-ethanolate of a compound are usually written in the same way. Professor Easton considered that this would be the case for aripiprazole. He said: 27. … Sometimes a chemical description may refer to the compound as being a particular hydrate or solvate, particularly if that conveys properties of the compound that are different to those of another hydrate or solvate of the same compound. However, I would normally expect an experiment to distinguish between different solvates of a compound only where the difference is relevant for the purposes of the experiment (and I consider that it is not here, given the subsequent steps in lines 16-42 of dissolving the product in ethanol as part of the synthesis of making the hydrochloride, sulfate, fumarate and maleate salts). 170 Thirdly, he would not regard ethanol in the crystals obtained from lines 5 to 15 of Example 1 to be an impurity. There would be no reason not to test the product in its solvate (including ethanolate) form. He said that some pharmaceuticals are in fact ethanolates. He exemplified darunavir ethanolate as a protease inhibitor used for treating auto-immune disease. He said that, as ethanol is of low toxicity, its removal is not essential for pharmaceutical applications. 171 He said that, if it had been desirable to avoid ethanol in the product, he would have expected lines 5 to 15 of Example 1 to use a solvent other than ethanol for the recrystallisation procedure.

[24]

The McGeary experiment 172 Associate Professor McGeary carried out an experiment in accordance with his protocol at the School of Chemistry and Molecular Biosciences at the University of Queensland. A number of people were present as observers, including Professor Easton. 173 Many of the details recorded by Associate Professor McGeary in his first affidavit concerning this experiment do not now call for comment. I will only refer to those matters that have some bearing on the issues that I now have to decide. 174 After performing the first recrystallisation of the crude aripiprazole in ethanol, Associate Professor McGeary observed a small amount of yellow material on top of the recovered product. He also observed "some instability in the balance" when weighing this product in that the weight of the sample decreased slightly as it was left lying on the balance. Associate Professor McGeary opined that the presence of the yellow material and the decrease in the weight of the sample indicated that impurities remained after the first recrystallisation step. After the second recrystallisation, the product was "washed well" with around 50 ml of ethanol and allowed to dry at suction for around 30 minutes. Associate Professor McGeary said that the resulting crystals were "cream and glistening". He said, however, that it was necessary to oven-dry the crystals in accordance with step 4.9 of the McGeary protocol. 175 At this stage, NMR spectral analysis of samples taken from these crystals showed that they were aripiprazole with residual ethanol present. Associate Professor McGeary said that, other than the presence of residual ethanol, the NMR spectra indicated that no significant impurities were present after the second recrystallisation. The spectra indicated that approximately one molecule of ethanol was present in the crystals for every two molecules of aripiprazole. This suggested to Associate Professor McGeary that ethanol was present in the crystals as a solvate (that is, it was bound within the crystal structure). 176 A melting point analysis after the second recrystallisation and 30 minutes of suction drying showed the melting point of a sample of these crystals to be 138.2°C to 139.2°C. Associate Professor McGeary observed that, when this analysis was performed, the crystals in the sample changed from "a glistening appearance" to "a chalky white colour". The analysis was repeated the next day. The recorded melting point on this occasion was 138.6°C to 139.3°C. 177 In order to remove the remaining ethanol from the crystals that had been dried for 30 minutes under suction, Associate Professor McGeary placed them in a vacuum desiccator and left them to dry overnight. Associate Professor McGeary said that the crystals "appeared physically unchanged" the next morning. Their weight was also "essentially unchanged". At the time, this suggested to Associate Professor McGeary that drying in the vacuum desiccator had not, and would not, remove the remaining ethanol. NMR spectral analysis showed, as it had the previous day, that approximately one molecule of ethanol was present in the crystals for every two molecules of aripiprazole. Again, other than the presence of residual ethanol, the NMR spectra indicated that no significant impurities were present in the crystals of the second recrystallisation. 178 Associate Professor McGeary concluded that the ethanol was present as a solvate - the crystals were aripiprazole ethanolate - and that vacuum drying at room temperature would not remove the residual ethanol. He decided that further drying should be carried out in accordance with step 4.9 of his protocol. He said that this is how he would have followed the directions in lines 5 to 15 of Example 1 of the 141 application as at September 2001. To this end, he took some of the crystals that had been stored overnight in the vacuum desiccator, placed them in a porcelain dish, covered the dish with a watch-glass, and placed the covered dish in an oven preheated to 92°C. The crystals were removed from the oven after 24 hours. At the time of removal of the crystals, the oven thermometer recorded the temperature to be 94°C. 179 Associate Professor McGeary observed the bulk of these oven-dried crystals to be white and crystalline (white crystals). He mixed the crystals to see whether the sample was homogenous. He observed that there was brown material on the bottom of the porcelain dish (brown crystals). These crystals were physically separate from the white crystals and had adhered to the bottom of the porcelain dish. 180 The total weight of the sample after drying was 19.20 g. The calculated weight loss on drying was 1.035 g. Associate Professor McGeary said that this weight loss was consistent with the loss of ethanol that would be expected if ethanol was present as a solvate (specifically, as an ethanol hemisolvate) in the substance before drying. NMR spectral analysis of the white crystals indicated that, after oven drying, that product was essentially pure aripiprazole. No ethanol was present. Upon further analysis, the white crystals were shown to have each of the features of claim 12 of the patent. 181 The brown crystals were also analysed. According to Associate Professor McGeary, NMR spectral analysis confirmed that the brown crystals were aripiprazole, with no significant impurities. That description of the brown crystals was a matter of some contention in the evidence. Associate Professor McGeary said that the brown crystals were slightly less pure than the white crystals, but no ethanol was visible in the NMR spectra. XRPD analysis indicated that the brown crystals and the white crystals had the same crystal form. The melting point of the brown crystals was slightly depressed and had a broader range than the melting point of the white crystals. IR spectroscopy also supported the conclusion that the brown crystals were aripiprazole. Associate Professor McGeary said that the overall similarity of the IR spectra obtained for the brown crystals and the white crystals confirmed that the two sets of crystals were the same substance. He reached the following conclusion about the brown crystals (in his first affidavit): 264 In my opinion, it is likely that the impurity present in the [brown crystals] is a very minor decomposition product that formed as a result of localised overheating of the crystals. The porcelain dish in which the crystals were heated in the oven was placed on the metal floor of the oven (which was covered by foil). Given the direct contact with the metal oven floor, it is likely that a portion of the base of the dish was at a higher temperature than the measured 92 - 94°C. Material in contact with that part of the base of the dish could therefore have been subjected to a higher temperature than the rest of the material. Whatever the cause, the effect was localised to crystals at the very base of the porcelain dish.

[25]

Other evidence concerning the McGeary experiment 182 Professor Black was not an observer at the McGeary experiment. He was provided, however, with an extract from Associate Professor McGeary's laboratory notebook in which Associate Professor McGeary made notes of the experiment. He was also provided with photographs taken during the course of the experiment. He expressed the general opinion that the McGeary experiment faithfully followed the McGeary protocol, "incorporating only steps not specified in the [McGeary protocol] that would have been routinely taken by an organic chemist in September 2001". 183 Professor Black said that the yellow material observed by Associate Professor McGeary after the first recrystallisation indicated that there was an impurity in the recrystallised product. He said that this indicated that a second recrystallisation was required to further purify the product. He also noted that Associate Professor McGeary had recorded instability in the balance during measurement of the sample weight. He said that this indicated that there was residual solvent in the product that was being lost during the measurement. Speaking as at September 2001, he said that he would have proceeded with a second recrystallisation if the product of the first recrystallisation was dry on visual inspection. He said that he would not have been concerned with removing all solvent at this stage. 184 Professor Black noted that the NMR spectral analysis of the product of the second recrystallisation confirmed the presence of aripiprazole and ethanol, as well as traces of the solvent used for the NMR spectral analysis. In his first affidavit, he said: 117. As the NMR analysis indicated the presence of residual recrystallisation solvent, ethanol, in the recrystallised product, as a standard practice I would have undertaken further steps to dry the product (ie, to remove the ethanol). In my opinion, the steps outlined in paragraph 4.9, which Associate Professor McGeary subsequently followed … were, in September 2001, conventional steps to follow to remove the remaining solvent. I know that organic chemists routinely used the same techniques to remove residual solvent in 2001. I would have routinely taken the same approach to remove the residual ethanol, and in my opinion, other organic chemists would have done the same … 185 Professor Black also noted that, after the product was dried in a vacuum desiccator, the NMR spectral analysis confirmed the continuing presence of residual ethanol. Speaking once again as at September 2001, he said that, as residual ethanol remained, he would have proceeded to oven-dry the product. 186 With respect to Associate Professor McGeary's observation of white and brown crystals, Professor Black noted that Example 1 states that the product of lines 5 to 15 is "colorless". Professor Black took "colorless" crystals to be synonymous with "white" crystals. He noted that Associate Professor McGeary had recorded the melting point of the white crystals to be 138.8°C to 139.9°C. Professor Black said that this range was consistent with the melting point of 139.0°C to 139.5°C reported in Example 1. 187 He noted that the NMR spectra obtained from the sample of the white crystals identified aripiprazole but not ethanol. Professor Black expressed the opinion that the NMR spectra, combined with the melting point analysis and visual appearance of the white crystals, indicated that they were the intended product of lines 5 to 15 of Example 1. Professor Black said that it was likely that the brown crystals observed by Associate Professor McGeary were the product of localised overheating during the drying process which resulted in the formation of an impurity affecting the specific crystals that had been exposed to the overheating. 188 Later in his first affidavit, Professor Black reviewed the physicochemical analyses performed on the white crystals in accordance with the McGeary protocol. He concluded that the analytical data for the white crystals described a substance identical to Crystals B. In this connection, it is to be noted that Professor Black relied on the TGA results to determine the hygroscopicity of the resultant crystals. 189 Dr Rowe also reviewed the physicochemical analyses performed on the white crystals and concluded that the crystals were Crystals B. 190 As I have noted, Professor Easton observed the McGeary experiment, at least up to that stage when the crystals were removed from the vacuum desiccator and placed in the oven for further drying in accordance with step 4.9 of the McGeary protocol. 191 Professor Easton advanced a number of criticisms of the steps taken by Associate Professor McGeary when carrying out the recrystallisation steps. In particular, he considered the steps taken by Associate Professor McGeary during the first and second recrystallisations to be inconsistent with what he considered to be the normal practice for organic chemists when they perform a recrystallisation. Nevertheless, he noted that the appearance of the crystals at the conclusion of the 30 minutes of drying at suction after the second recrystallisation was "cream and glistening". He said that he regarded that description to be the same as "colorless" as used in Example 1. He also noted that the two melting point analyses undertaken by Associate Professor McGeary showed the crystals to have a melting point represented by the following ranges: 138.2°C to 139.2°C and 138.6°C to 139.3°C. 192 Professor Easton continued (in his second affidavit): 49. … I also noted that the crystals at this stage were flakes, by which I mean small flat pieces. I consider the melting points referred to … to be what I and other organic chemists would consider to be within the range of 139.0-139.5 °C, because they are well within normal experimental variation. Accordingly, I consider the product obtained at this stage to be colourless flake crystals with a melting point of 139.0-139.5 °C as is required by lines 14-15 of Example 1 and there was no need to perform any further steps, as the product of lines 5-15 of Example 1 had already been reproduced. 193 Professor Easton then commented on the further drying step undertaken by Associate Professor McGeary. Professor Easton said that the product has known sensitivity to heat which should have precluded oven drying. He said that Associate Professor McGeary observed a change in the crystals during the melting point analysis from a glistening appearance to a chalky-white appearance. He said that this change indicated to him, and that other organic chemists would also understand, that the product was sensitive to heat. 194 He remarked that, if the ethanol present was in fact ethanol recrystallisation solvent, and not solvate ethanol (so that the product was not an ethanolate), he would have used techniques known to him, and other organic chemists, at September 2001, for removing ethanol recrystallisation solvent. 195 Professor Easton said that if, contrary to his own view, the product of lines 5 to 15 of Example 1 should not be an ethanolate, and all ethanol should be removed, he would not have chosen (and he believed other organic chemists would not have chosen) the methods which Associate Professor McGeary used. In this connection, he said that "[t]he obvious way" was to recrystallise the material in the absence of ethanol (that is, by using another solvent). He accepted in cross-examination, however, that such a step would represent a departure from Example 1. 196 In any event, Professor Easton questioned why Associate Professor McGeary chose to remove ethanol by heating the material at 80°C to 100°C for 24 hours. Professor Easton suggested that that step was contrary to the cautionary remarks made in Vogel's Textbook of Practical Organic Chemistry (Vogel), a reference text to which Associate Professor McGeary referred in his own evidence, and a text which Professor Easton accepted was one which he and other organic chemists consulted in September 2001 when difficulties arose during experiments, including in relation to recrystallisation and the drying of a solid form. 197 Professor Easton noted that, at atmospheric pressure, the boiling point of ethanol is about 78°C. At lower pressures (that is, under vacuum), it is reduced to less than normal room temperature. He said that Associate Professor McGeary had no reason to heat the product above the boiling point of ethanol because, given the vapour pressure of ethanol, heating above boiling point would not increase the volatility of the ethanol in the crystals. Professor Easton explained the position as follows: 67. In fact the boiling point of ethanol is irrelevant because if one wished to remove the ethanol (that is, the solvate ethanol rather than the ethanol recrystallisation solvent) from the crystal lattice, the temperature is that required to disrupt the ethanol interaction with the crystal structure, not ethanol-ethanol interactions. In other words, the boiling point of ethanol is determined by the heat required to dissociate an ethanol molecule from other ethanol molecules, which is unrelated to the temperature needed to remove solvate ethanol which is determined by the heat required to dissociate an ethanol molecule from the crystal lattice. 68. It is not necessary for a solvent to boil to evaporate if, like ethanol, it has a relatively high vapour pressure below its boiling point. Even if it is necessary to boil the solvent (which I do not accept as correct), that is no justification for heating the product at a temperature range significantly above the boiling point. Although the McGeary [p]rotocol at step 4.9 states that the product is to be heated between 80 °C to 100 °C, I note that Associate Professor McGeary in fact commenced heating of the product at 92 °C. I am also informed by page 16 of Associate Professor McGeary's notebook … that the temperature of the oven at the time the product was taken out was 94 °C. Associate Professor McGeary did not explain why these temperatures, which are in the upper end of the 80 °C to 100 °C range, had been chosen as opposed to temperatures in the lower end of the range. 198 Professor Easton noted that there was clearly brown material within the sample after this heating step. He said that the brown crystals were not a successful reproduction of the instructions in lines 5 to 15 of Example 1, which require "colorless" crystals. He considered the brown crystals to represent decomposition, which might be a more extreme case of the yellowing that Associate Professor McGeary had observed during recrystallisation. 199 Professor Easton reviewed the analytical results for the white crystals. He observed that there was nothing in the 141 application that would enable him to determine the XRPD spectrum, IR spectrum, endothermic peak in TGA and DSC or water content of the crystals obtained at the end of lines 5 to 15 of Example 1. He could only rely, therefore, on the melting point and appearance of the product to determine whether the crystals met the description of the product that should have been obtained. He agreed that the white crystals obtained by Associate Professor McGeary matched that description. However, he stressed that such crystals had already been obtained at a much earlier point in the McGeary experiment. 200 With respect to the hygroscopicity of the white crystals, Professor Easton noted Professor Black's and Dr Rowe's reliance on TGA to determine that the white crystals were not hygroscopic under the test conditions for Crystals B. Professor Easton said that the results of the TGA analyses that were performed were presented on different scales and were difficult to compare directly. He also said that his own inspection showed the response profiles to be different, which indicated to him that the material had changed. He said that only the Karl Fischer method unambiguously determines the water content of a sample. 201 Despite the reservations expressed by Professor Easton, I am satisfied that the white crystals obtained by Associate Professor McGeary as a result of his experiment were Crystals B.

[26]

The White protocol 202 The applicants' solicitors approached Associate Professor White to prepare a written protocol detailing how, as at 25 September 2001, he would perform the procedure described in lines 5 to 15 of Example 1 of the 141 application (the White protocol). 203 The White protocol is divided into Parts A and B. Part A concerns the conversion of the hydrochloride salt of 1-(2,3-dichlorophenyl)-piperazine into its free base form. Part B concerns the preparation of 7-{4-[4-(2,3-dichlorophenyl)-1-piperazinyl]butoxy}-3,4-dihydrocarbostyril. Relevantly, it involves steps 1 to 24. 204 In his affidavit, Associate Professor White said: 44. I consider the chemical reactions envisaged by steps (1) to (24) of Part B of the Protocol to be routine and straightforward. I consider the steps contained in steps (1) to (24) of Part B of the Protocol to be what I and other organic chemists would have done as at 25 September 2001 to carry out such reactions. 205 Associate Professor White made the following observations regarding the recrystallisation step in lines 5 to 15 of Example 1 of the 141 application: 67. Recrystallisation is a process by which a crude crystalline material is dissolved in a solvent which is better at dissolving the material at a higher temperature as compared to a lower temperature, so that as the solution cools down it becomes supersaturated and crystals are re-grown. These crystals would be more pure than the crude material, as any impurities which were in the crude would have been dissolved and left behind in the solution. Therefore, when I and other organic chemists read 'the residue thus obtianed [sic] was recrystallized from ethanol twice' in the fourth sentence (lines 11-12) of Example 1, I and other [organic] chemists would understand this to mean: (a) First, the material would be fully dissolved in hot ethanol. It is desirable to use ethanol which is at or within 1-2°C of its boiling point because this avoids adding too much ethanol to dissolve the solid which may reduce the recovery of the product after crystallisation. (b) Then, the solution would be cooled down to allow more pure crystals to be grown in the solution. (c) The crystals are collected once they are dried of excess solvent, ie removed of such solvent as one would be able to remove using standard methods and within a reasonable timeframe, which is generally solvent adhered to the surface of and therefore in excess to, the solid matter. (d) The entire process is repeated a second time, with the crystals collected when constant weight has been attained. 206 He also said: 71. While a second recrystallisation would be a repeat of the first recrystallisation, there would be some necessary differences including: (a) how much solvent to use, because generally the more pure the crystals are, the more solvent would need to be used in further removing any impurities; and (b) how much effort to use in removing the excess solvent, ie with a final recrystallisation, it would be more important to ensure that the product is as dry of excess solvent as possible by the use of standard methods and within a reasonable timeframe, until constant weight has been attained. 207 In his protocol, Associate Professor White proposed that, when the second recrystallisation step appeared to be nearly complete, the mixture would be chilled on an ice bath for 20 minutes (step 22). In step 23: (23) The solid in the flask is broken up with a spatula and vacuum filtered using a clean porosity-3 or 4 scintered glass funnel with 5.5 cm internal diameter and 10 cm height. The solid is packed down with the back of a spatula while applying vacuum. The vacuum is released, the solid is wetted with ice-cold ethanol (20 ml), then suction is immediately applied until no more solvent is filtering through. The solid is left on the vacuum filtration apparatus for 30 minutes after the ethanol has filtered through and the filter funnel is no longer cold to the touch. 208 In his affidavit, Associate Professor White said that, based on his experience, he would expect that vacuum filtration under step 23 would remove at least 95% of the excess solvent. He said that, in order to remove the remaining 5% excess solvent, it would be appropriate to place the material in a vacuum oven under the conditions stated in step 24. Step 24 is as follows: (24) After removal of the boiling chips the solid is transferred to a pre-weighed crystallizing dish (9.5 cm x 5.5 cm), and the combined weight of the solid and the crystallizing dish is measured on a Sartorius top-loading balance. The sample is then covered with aluminium foil and placed in a vacuum oven. The oven is heated to 40°C and a vacuum of 15-20 mm Hg is applied. After 1 hour the sample is removed from the oven and the weight is measured, then the sample is returned to the oven, and the vacuum re-established. After a further hour of drying the sample is removed from the oven and weighed. If the weight is unchanged, then the sample is deemed to be dry and is transferred to a chemical storage jar. If the sample is not yet constant weight then the above process is repeated until the sample has reached constant weight.

[27]

Other evidence concerning the White protocol 209 Professor Easton said that, while there are some steps in the White protocol that he would have done differently, none of the differences would have had any material effect on the product obtained by following that protocol. In his second affidavit, he said: 79. … I consider that the White protocol fundamentally represents how I, and other organic chemists, would have reproduced lines 5-15 of Example 1. 210 With respect to step 24, Professor Easton said that he would have tried one or more of the following methods of drying the product: Gently agitating the filter cake, for example, with a spatula, to ensure that it was properly aerated. Spreading out the material as a thin layer on a Petri dish and leaving it at room temperature to allow the last traces of solvent to evaporate. Using a vacuum under air-flow to pull the solvent out. Using an alternative, more volatile solvent to wash or recrystallise the product. 211 He said that his preference would have been to leave the product to dry in a Petri dish covered with filter paper, until constant weight had been reached. He considered this to be a more conventional method than the one proposed in the White protocol. He said that he would only use a vacuum oven as a last resort, where the material would be gently warmed at no higher than 50°C for short periods. Nevertheless, he said that drying as proposed in the White protocol would not have materially affected the product obtained. It would simply accelerate the drying process, without risking a change in the form of the product. 212 Professor Black made a number of comments on the White protocol. I will confine my summary to his comments with respect to step 24, the drying step proposed with respect to the solid material obtained from the second recrystallisation after suction filtration. 213 Professor Black said that it was inherent in the direction in lines 5 to 15 of Example 1 to recrystallise the residue "from ethanol twice" that any recrystallisation solvent remaining in the product after the second recrystallisation should be removed. He observed that the White protocol provided that the material would be placed in a vacuum oven and heated at 40°C. He said that, in September 2001, it was, in his experience, more common to dry material (to remove solvent) in a vacuum desiccator at room temperature, rather than in a vacuum oven. He said that if, after drying in this manner, NMR spectral analysis indicated that ethanol remained, he would have considered using a conventional oven or a drying pistol. He said that he had never considered a vacuum oven to be a standard piece of laboratory equipment. He remarked that his own research group does not have a vacuum oven and that the School of Chemistry at the University of New South Wales owns only one such oven, but has at least 12 conventional ovens. 214 Perhaps more importantly, in his second affidavit, Professor Black said: 29. Paragraph (24) also provides for the weight of the material to be measured until constant weight is reached. In September 2001, it was not common to determine dryness of a product after a final recrystallisation by achieving constant weight, but even if so, it would have been unusual to measure the weight in one hour intervals. In my experience, it would have been more common to leave the material to dry overnight or for 24 hours and to determine whether the material was dry using NMR analysis or elemental analysis. The question of whether a material is appropriately dry depends upon the purpose for which the recrystallised material will be used. That is, the extent of drying required depends upon the context in which the recrystallisation is being performed, particularly the use of the recrystallised product. … [A]nalytically pure aripiprazole should be obtained from the method described in Example 1. 215 He said that, if NMR spectral analysis indicated that ethanol, as the recrystallisation solvent, remained present, he would undertake further drying of the material "as an inherent part of the recrystallisation step in Example 1".

[28]

The White experiment 216 Associate Professor White carried out an experiment in accordance with his protocol. The steps in Part A of his protocol were carried out in advance of the steps in Part B. A number of observers, including Professor Black, were present when the Part B steps were carried out. 217 As with the experiment carried out by Associate Professor McGeary based on the McGeary protocol, many of the details recorded by Associate Professor White in his affidavit in relation to the carrying out of the experiment in accordance with the White protocol do not now call for comment. It is only necessary for me to refer to the following matters. 218 The relevant steps in Part B were carried out over two days, with the two recrystallisation steps being carried out on the second day. 219 With respect to step 24, it was not necessary for Associate Professor White to remove the boiling chips because, as matters transpired, they had been removed during the course of Associate Professor White carrying out the second recrystallisation step. Nothing turns on that change. However, when seeking to dry the solid material in the vacuum oven, Associate Professor White was not able to establish the pressure condition of 15 to 20 mm Hg as stipulated in his protocol. The lowest pressure condition he could establish on that day was 27 to 30 mm Hg. The only effect of this appears to have been that, at this higher pressure, drying could take longer. The weight of the Petri dish together with the solid product just before it was placed in the vacuum oven was 131.994 g. The weight of the solid material placed in the oven was 39.267 g (the Petri dish weighing 92.727 g). 220 After approximately one hour and 25 minutes drying, Associate Professor White observed that some weight had been lost: the Petri dish and solid product weighed 131.750 g. After approximately 45 minutes of further drying under the same conditions, Associate Professor White observed further loss of weight: the Petri dish and solid product weighed 131.746 g. Associate Professor White concluded that constant weight had not then been attained. Associate Professor White then proceeded as follows: 132. The product was placed back in the vacuum oven under the same pressure conditions for a further 25 minutes. I decided to check the product at this point in time (instead of waiting for another hour as stated in the [White protocol]) because the product had already been close to constant weight the last time I took it out of the vacuum oven and I did not wish to use any more time than that was necessary to dry the product to constant weight. The weight of the Petri dish together with the product was determined to be 131.746 g and I concluded that constant weight had been attained. … 221 Associate Professor White described the appearance of the product as "colourless crystalline flakes". He transferred the product to a chemical storage jar with a lid and placed it in a portable desiccator as he considered this to be "an appropriate storage method". 222 Associate Professor White described the final yield from his experiment as follows: 134. The final yield of the product was 39.019 g. While this is lower than the 57 g reported in Example 1, I was not concerned because problems caused by the insoluble material during the extraction stage would have accounted for some of the yield. Further, a certain portion of the loss in yield can be attributed to those crystals which were still growing in the filtrate which had been vacuum filtered off during my performance of step (23). Now shown to me … is a photo of the filtrate showing the presence of these further crystals. On 12 July 2012, I vacuum filtered this filtrate so as to estimate approximately how much product had been lost. Now shown to me … is a photo of the solid material obtained after vacuum filtration. I scraped this material out and determined that it weighed 1.9 g. 223 Over the following days, this product was analysed. In his affidavit, Associate Professor White said that NMR spectroscopy showed that the product of step 24 had 0.2 equivalents of ethanol. This is a quantity of ethanol less than expected for a hemiethanolate. 224 Associate Professor White's evidence about how humidity tests were carried out on this product (60°C and 100% relative humidity over 24 hours) led to some debate between the experts, at least in the preparation of their evidence, as to whether the results obtained from these tests on Associate Professor White's final product were reliable. However, as senior counsel for Apotex very properly volunteered on the first day of the hearing, it is not necessary to resolve that particular issue because the XRPD analysis carried out on Associate Professor White's final product showed that it did not possess, in any event, the XRPD spectrum characteristic of Crystals B as claimed in claim 12 of the patent. As senior counsel correctly observed, the real issue between the parties is whether, in carrying out the directions of lines 5 to 15 of Example 1 of the 141 application before 25 September 2001, the person skilled in the art would have effectively stopped at step 24 of the White protocol, or gone further to seek to remove ethanol as Associate Professor McGeary sought to do by carrying out step 4.9 of his protocol. Nevertheless, the Karl Fischer analyses carried out on Associate Professor White's final product showed that, when stored at 60°C with 100% relative humidity for 24 hours, it had a water content in excess of that for Crystals B as claimed in claim 12. 225 With respect to the fact that the crystals resulting from his experiment contained ethanol when dried to weight, Associate Professor White gave this evidence: 149(a) The fact that the product of step (24) contains ethanol does not change my conclusion. The isolation of a solid as a solvate (ie containing solvent of crystallisation) is not unusual to me. In my experience as a crystallographer I have frequently encountered crystalline materials containing solvent molecules as part of the crystal lattice (known as solvates). In the present example I am also not surprised as ethanol is a hydrogen bond donor and can form relatively strong hydrogen bonds to substances containing amine hydrogen bond acceptor groups, which is the case with [aripiprazole] and this may lead to retention of the solvent as part of the crystal lattice. My conclusion is also not affected by the fact that the product may be used as a pharmaceutical product. 226 He also gave this evidence: 149(b) Further, as I understand, in making each of the salts referred to in the remainder of Example 1, the product of lines 5-15 of Example 1 is to be redissolved in or recrystallised from ethanol, so that it is immaterial that the product of lines 5-15 of Example 1 contains ethanol. 227 Associate Professor White said that, regardless of whether he had known that the product to be obtained was a solvate or non-solvate, and regardless of whether the product was to be used to make one of the salts referred to in the remainder of Example 1, he would not have written his protocol differently or conducted his experiment differently. 228 In cross-examination, Associate Professor White accepted that his evidence quoted in [226] above was based on his understanding that the product obtained as a result of following the directions in lines 5 to 15 of Example 1 would be recrystallised in ethanol for the purpose of making the salts referred to in the remainder of the example. His attention was directed in cross-examination to the pharmaceutical composition exemplified in the 141 application as suitable for a tablet formulation. He accepted that this composition used aripiprazole in its free base form. He said that, if the product of lines 5 to 15 of Example 1 were to be used in its free base form, one would need to know if the product was a solvate in order to determine how much of the product should be put in the composition. In this connection, as I have noted, the 141 application discloses that 5 mg of aripiprazole should be used. Associate Professor White accepted that 5 mg of aripiprazole hemiethanolate would contain less than 5 mg of anhydrous aripiprazole. He also accepted that there was no suggestion in the 141 application that one might need to make adjustments in quantity in order to use the product of lines 5 to 15 of Example 1 to achieve the 5 mg of aripiprazole referred to in the composition. 229 Furthermore, in the course of Associate Professor White's oral evidence, it emerged that, before carrying out Part B of his protocol in the observed experiment, he conducted a "dry run". The product he obtained was a hemiethanolate. By conducting a dry run, Associate Professor White was able to refine his protocol by making some changes. His evidence was, however, that these changes were slight and related to matters such as the volume of materials to be used. 230 Another consequence of conducting a dry run was that, after conducting the observed experiment, Associate Professor White realised that the product of the dry run and the product of the observed experiment were different. He observed that the crystals from the observed experiment appeared to be a mixture. He gave this evidence: What was the different conclusion that you came to and how did you come to it?---I just - I had trouble reconciling the dry run results with this result where I pumped to constant weight, and there were a couple of differences between the dry run and this one. One was that in the dry run the crystals had grown a lot longer and they were - they appeared to be larger, and in this run the crystals appeared on average to be smaller when you just viewed them. Therefore I decided that perhaps we had formed maybe a solvate, or maybe a mixture of solvates and non-solvates. So I examined the material under the microscope in our x-ray laboratory and I chose a large crystal which looked very similar to the crystal that I had looked at from the dry run, and I completed an analysis of it and I found it to be a hemiethanolate which was identical to the previous experiment. I then looked under the microscope and found there were large crystals and there were actually small crystals as well which had the same appearance; they were plates and sort of off white. So I decided to do a full analysis I would also look at the crystal structure of that. So I determined the structure of that crystal as well and I determined it to be a non-solvate. So it appeared the second run had formed a mixture of two forms of aripiprazole, one was a solvate and one was a non-solvate.

[29]

Other evidence concerning the White experiment 231 Professor Black observed the White experiment as it related to the carrying out of Part B of the White protocol. Professor Black made a number of criticisms of the way in which the experiment was conducted. It is only necessary, however, for me to summarise Professor Black's evidence so far as it concerns the carrying out of step 24 of the White protocol. 232 As part of carrying out this step, Associate Professor White informed Professor Black that the vacuum oven had been equilibrated at 41°C. Associate Professor White showed Professor Black a thermometer which recorded this temperature. However, according to Professor Black, the thermometer was not left in the vacuum oven while the material was drying. Moreover, it does not appear that a data logger or other device was used to confirm the temperature in the oven during the drying period. 233 There is evidence to suggest that the oven was, in fact, set at 60°C. However, in response to a query raised by Professor Black, Associate Professor White confirmed that, while the oven was set at this temperature, the temperature inside the oven was measured by Associate Professor White using the thermometer. It is unclear why the temperature inside the vacuum oven would have been different from the temperature to which it was set. Although this discrepancy was canvassed in the evidence, ultimately nothing seems to have turned on it. 234 As I have noted, Associate Professor White concluded on the second weighing of the Petri dish and material at 131.746 g that constant weight had been achieved, and that the product of the White protocol had been attained. However, Professor Black said that, on the occasion that Associate Professor White weighed the Petri dish and product at 131.746 g for a second time, Professor Black in fact observed the weight of the material to fluctuate and, on balance, show an overall increase in weight. Professor Black said that, if he had been conducting the experiment, he would have left the material to dry overnight in the vacuum oven after the first weighing at 131.746 g. In his second affidavit, he said: 64. … I would also have considered increasing the temperature of the vacuum oven to 50°C or 60°C. In my opinion, if the material was left in the vacuum oven overnight, I could confidently conclude that all the ethanol that could be removed by drying in the vacuum oven had been removed. 235 Professor Black said that the subsequent NMR spectral analysis showed that the drying process had not removed all the ethanol and that the final product should have been further dried to remove it. 236 Overall, Professor Black concluded: 67. The deviations taken from the White [p]rotocol that I observed may have impacted on the yield of aripiprazole but they did not prevent the production of aripiprazole. The presence of aripiprazole in the material produced by Associate Professor White was confirmed by NMR analysis … However, the NMR analysis of that material also indicates that recrystallisation solvent, ethanol, remained after drying in the vacuum oven. The presence of detectable ethanol indicates that the material obtained required further drying, as the product of Example 1 should be analytically pure aripiprazole … 237 Based on the results of the NMR spectral analysis of the final product, Professor Black calculated that there were approximately 0.14 molecules of ethanol to each molecule of aripiprazole. On the other hand, as I have noted, Associate Professor White calculated that 0.2 equivalents of ethanol were present. 238 Professor Black accepted that the measured melting point of 138.2°C to 139.0°C of the product obtained by Associate Professor White was consistent with the melting point of the product recorded in lines 5 to 15 of Example 1 of the 141 application. However, Professor Black made the following observation in his second affidavit: 126. … I know from the NMR analysis that Associate Professor White's Final Product contained ethanol. Melting point analysis involves heating a sample whilst observing it to determine the temperature at which it melts. Ethanol present in the sample being analysed would have been removed from the sample during heating and prior to it melting. This means that the Final Product would have been dried of remaining ethanol prior to the melting point being determined, and the melting point determined is the melting point of a dried (of ethanol) substance, rather than the Final Product. 239 Professor Black criticised Associate Professor White's conclusion that, based on the appearance and melting point of the final product, he had successfully reproduced the directions in lines 5 to 15 of Example 1. First, Professor Black said that this conclusion could not be drawn simply from the appearance of the final product and its melting point. Professor Black said that a successful reproduction of the directions in lines 5 to 15 of Example 1 could not be claimed in the absence of NMR spectral analysis indicating whether the crystals were aripiprazole and that ethanol had been removed from the crystals obtained. In this connection, he referred again to the fact that the melting point which Associate Professor White had actually determined was the melting point of the sample free of ethanol. Associate Professor White accepted in cross-examination that determining the melting point of the crystals he obtained might be inconclusive. He gave this evidence: You accept, don't you, that it is quite possible that in testing the melting point of a hemiethanolate that the ethanol would evaporate before the melting point of aripiprazole. You couldn't exclude that possibility, could you?---No. No. And so if in terms of a particular use of the aripiprazole crystals you had obtained in example 1, namely their free base use for the particular example pharmaceutical composition provided, if the NMR told you that you had residual solvent present, then the melting point was an inconclusive test as to whether you had done enough, wasn't it?---In hindsight that may be the case, yes. 240 Secondly, Professor Black argued that Associate Professor White seemed to imply that his final product was an ethanol solvate. He said, however, that Associate Professor White had not provided any explanation for the conclusion that the ethanol remaining in the final product was present as a solvate. In this connection, he observed that the solvent in solvates is present in simple ratio to the solvated compound. Thus, a ratio of 0.14 molecules of ethanol to each molecule of aripiprazole (that he had calculated) or a ratio of 0.2 molecules of ethanol to each molecule of aripiprazole (that Associate Professor White had calculated) was not consistent with the final product being an ethanol solvate. These ratios were, however, consistent with the product being a mixture that contained some ethanol solvate. 241 Thirdly, Professor Black argued that, in the context of the 141 application, the product of lines 5 to 15 of Example 1 is aripiprazole in a non-solvated, free base form. In this connection, he said that all the work using aripiprazole reported in the 141 application was based on aripiprazole being pure aripiprazole, devoid of residual solvent. He pointed to the 141 application reporting on tests of aripiprazole for its biological activity and said that the validity of this testing depended on knowledge of the molecular weight of the test compound, and the weight would be different if the compound was an ethanolate. 242 Professor Easton did not observe the White experiment. He did, however, review the results of the analyses performed on the final product obtained by Associate Professor White, as well as an image taken of that product. He agreed that the product had the appearance of "colorless crystalline flakes". He also agreed that the XRPD pattern of the product was quite different to that shown in Figure 5 of the complete specification. He concluded that this product had a different crystalline form to that of Crystals B claimed in claim 12 of the patent. He said that this was supported by the TGA results that showed a major endothermic peak at 108°C but only a minor peak near 141.5°C. He also noted that the Karl Fischer analysis of the product showed that it had, on average, a moisture content of 0.8%, well above the 0.40% level stated in claim 12 of the patent. 243 I am satisfied that the product obtained by Associate Professor White as a result of his experiment was not Crystals B.

[30]

Consideration 244 Apotex advanced two major submissions on this aspect of its case. The first submission was that it is not open to the applicants to contend that a reproduction of the directions in lines 5 to 15 of Example 1 produced anything other than anhydrous aripiprazole crystals free of ethanol. The second submission, advanced as an alternative, was that, as a matter of construction, the 141 application discloses the preparation of anhydrous crystals of aripiprazole from which the person skilled in the art would inevitably remove residual ethanol because aripiprazole in its free base form is disclosed as having been used for testing and for the preparation of a pharmaceutical composition. 245 Apotex sought to support its first submission by arguing that one of the methods of preparing anhydrous aripiprazole crystals described in the complete specification of the patent replicates lines 5 to 15 of Example 1 of the 141 application. It argued that there is no teaching in the complete specification that the method produces an ethanolate or requires a step to remove ethanol. It argued that "[i]t follows" that the person skilled in the art would remove residual ethanol as a matter of course. It said that the applicants "cannot impeach the teaching" of the complete specification. 246 I do not accept this submission or the reasoning on which it is based. The issue is whether the directions of lines 5 to 15 of Example 1 are an anticipatory disclosure of the invention as claimed. The resolution of that issue is not advanced by removing lines 5 to 15 of Example 1 from the context of the 141 application and thereafter viewing them in the context of the complete specification of the patent. It is not to the point that there is no teaching in the complete specification that an ethanolate is the product of following the relevant directions or that no step to remove ethanol is given. No issue of the applicants "impeach[ing] the teaching" of the complete specification arises. Indeed, it is difficult to see how it could arise. The described process in the complete specification on which Apotex relied is given as one for preparing conventional anhydrous aripiprazole crystals - that is, the crystals having the problem of hygroscopicity that the invention seeks to redress. 247 In support of its second submission, Apotex argued that, if residual ethanol is not removed as part of following the directions of lines 5 to 15 of Example 1, a greater quantity of aripiprazole would be needed to arrive at 5 mg of the desired active ingredient specified in the tablet formulation that is exemplified. This argument is certainly supported by the evidence given by Associate Professor White in cross-examination. Apotex also argued, correctly, that there is no indication on the face of the 141 application of the need for a calculation and consequent adjustment in the quantity of the material required to prepare the exemplified tablet formulation. Moreover, Apotex relied on the fact that the 141 application does not refer to the product of lines 5 to 15 of Example 1 as a solvate, such as a hemiethanolate. 248 Relatedly, Apotex argued that it is "utterly irrelevant that different scientists may choose different techniques or conditions to remove the ethanol". It put the issue this way: … The critical question of construction is whether the skilled addressee would understand the teaching to require [the removal of ethanol]. No witness called by the applicants has said that he would not have been able to remove the ethanol had he wanted to do so. Further, the law of novelty does not require the Court to "subject the repetition of prior art examples to a tyranny of precision which has little to do with the issue before the court or the science to which the prior art relates." What must be proved is "on the balance of probabilities, what would be the result of carrying out the example in the prior experiment." [Footnotes omitted] 249 This aspect of Apotex's second submission alludes to remarks made by Laddie J in Evans Medical Ltd's Patent [1998] RPC 517. Those remarks were made in respect of evidence of an experiment carried out to follow a method disclosed in an earlier patent application (the Takeda patent application) for making a B. pertussis product for development into a whooping cough vaccine. It was alleged that the Takeda patent application anticipated the invention as claimed in the patent in suit, which was concerned with antigenic preparations for use in acellular vaccines against whooping cough. 250 His Lordship's remarks were directed to the fidelity with which directions in a prior publication should be carried out in order for the result to be considered as an anticipation of an invention as claimed. This can be seen from the following passage in his Lordship's reasons (at 571-572): … One of the issues before the court is whether the carrying out of a specific example described in the prior art falls within the scope of the patent in suit. Normally the easiest way of demonstrating that it does is for the attacker to carry out that example precisely to see what results are obtained. If he makes alterations to the experimental protocol the court may well conclude that the alterations made a significant difference to the results obtained. If that is so, then the attacker will not have proved what happens when the described example is carried out. But the fact that this is normally the easiest way of proving what the example produces does not mean that it is the only way. For example a piece of prior art may include a reference to carrying out an experiment in a conical flask made by a particular manufacturer. By the time of the repetition that manufacturer may not exist. Carrying out the experiment in a flask from an alternative manufacturer would normally not be expected to have any impact on the course of the reaction. The experiment still proves, on a balance of probabilities, what would be the result of carrying out the example in the prior document. To approach this issue in any other way would be to subject the repetition of prior art examples to a tyranny of precision which has little to do with the issue before the court or the science to which the prior art relates. … 251 The issue raised by Apotex's second submission is not quite the same as that to which Laddie J's remarks were directed. The issue in the present case is not how given directions for removing ethanol should be carried out but whether, in carrying out the directions in lines 5 to 15 of Example 1 before the priority date, the person skilled in the art would have proceeded to remove ethanol from the resulting product, either as a matter of course or because of some teaching, recommendation or suggestion in the 141 application: Bristol-Myers Squibb Company v F H Faulding & Co Limited (2000) 97 FCR 524 at [67]. 252 On the evidence before me, I am satisfied that, when carrying out the directions in lines 5 to 15 of Example 1 of the 141 application, the person skilled in the art would have sought to dry the product of the second recrystallisation in an endeavour to obtain the colourless flake crystals. Associate Professor McGeary and Associate Professor White provided a drying step in their respective protocols. However, their proposals in that regard differed significantly. 253 In this connection, it is to be borne in mind that, as a result of following his protocol, and drying his crystals to constant weight, Associate Professor McGeary obtained aripiprazole crystals as a hemiethanolate. When Associate Professor White conducted his dry run based on his then draft protocol, he obtained crystals as a hemiethanolate. When he carried out his observed experiment, he obtained aripiprazole crystals that were a mixture of a hemiethanolate and a non-solvate. In each case, Associate Professor White's resultant product, after drying to constant weight, was therefore one containing ethanol. On the basis of this body of evidence, I find that, by following the directions in lines 5 to 15 of Example 1, and drying the crystals to constant weight, which the person skilled in the art would, at least, do, the resultant product is crystals of aripiprazole containing ethanol, most likely as a hemiethanolate. 254 It is convenient to focus, at this stage, on the matters that informed Associate Professor White as to the sufficiency of simply drying the resulting crystals to constant weight and taking no further step. 255 The substance of Associate Professor White's evidence was that obtaining a solvate was unsurprising, that a solvate could be used as a pharmaceutical product, and that Example 1 provided instructions for the formation of salt forms of aripiprazole using ethanol, meaning that it was immaterial that the product of lines 5 to 15 of Example 1 contained ethanol. Similar evidence was given by Professor Easton: see [167] to [170] above. 256 I accept that, if the person skilled in the art were to carry out the directions in lines 5 to 15 of Example 1 to obtain aripiprazole as an intermediate to be used to make one of the salt forms according to the directions in the remainder of the example, then that person may not be motivated to remove ethanol from that intermediate, beyond drying it to constant weight. But the directions in lines 5 to 15 of Example 1 must be considered in the context of the 141 application. The 141 application shows that aripiprazole can be used pharmacologically in its free base form. Indeed, the 141 application recommends or, at the very least, suggests that the product of lines 5 to 15 of Example 1 be so used. Reference need only be made to the disclosure of the pharmacological testing of aripiprazole in that form and to the exemplified pharmaceutical composition for the tablet formulation. It is true that the 141 application speaks, in general terms, of the use of ethanol - as well, it should be said, of the use of a great many other substances - as a binder or carrier in pharmaceutical compositions containing aripiprazole (and, indeed, containing other carbostyril derivatives). That indication does not detract, however, from the plain disclosure that at least one intended purpose of carrying out the directions of lines 5 to 15 of Example 1 is to obtain anhydrous aripiprazole crystals in their free base form. 257 I am satisfied that the person skilled in the art might well be motivated to carry out the directions in lines 5 to 15 of Example 1 with that end in view and, if so, would take steps, as a matter of course, to remove ethanol from the crystal lattice. That would have been the case before 25 September 2001, as it would be the case now. Thus, if that end were sought to be attained, I do not accept that the White experiment went far enough. It did not involve all the steps that a person skilled in the art would carry out in following the teaching, recommendation or suggestion to use, and hence obtain, anhydrous aripiprazole crystals in their free base form. 258 It will be apparent from this conclusion that I do not accept, as Associate Professor White and Professor Easton seemed to suggest, that the 141 application directs or perhaps suggests that one form of aripiprazole should be obtained or used to the exclusion of another. Rather, it discloses that, by following the directions of lines 5 to 15 of Example 1, the person skilled in the art would obtain aripiprazole which he or she can choose to use in its free base form or in one of the disclosed salt forms. The 141 application discloses that both forms of aripiprazole (that is, the free base form or one of the disclosed salt forms) can be used. In so doing, it makes clear that the path to obtaining both forms includes carrying out the directions in lines 5 to 15 of Example 1. If the person skilled in the art were motivated to obtain anhydrous aripiprazole crystals in their free base form, the parties did not suggest that, before 25 September 2001, the person skilled in the art could not or would not take routine steps to achieve that end. 259 The evidence shows that such steps would involve the further drying of the product of the second recrystallisation (that is, not merely to constant weight). However, the choices that could be and could have been made in that regard assume considerable importance when determining whether the directions of lines 5 to 15 of Example 1 are themselves anticipatory in the required sense. Neither Example 1 nor any other part of the 141 application gives specific direction as to how any such step or steps could or should be undertaken. Indeed, the 141 application can be said to be indifferent in this regard. I now turn to consider those choices. 260 By step 4.9 of his protocol, Associate Professor McGeary proposed that further drying would be undertaken in an oven at between 80°C and 100°C for an initial overnight period (or equivalent). Associate Professor McGeary explained his choice of this range. He said that he selected 80°C as this was around the boiling point of ethanol (approximately 78°C). He said that he selected 100°C as that was, in his opinion, a "reasonable choice" because that temperature was "below the melting point of aripiprazole" (approximately 139°C) and therefore unlikely to degrade the crystals. 261 As summarised above, Professor Easton was critical of this step: see [193] to [197] above. Relevantly for present purposes, he referred, first, to the teachings in Vogel. Vogel teaches that, for low melting point solids, the best method of drying is to place the crystals in a watch-glass in a dessicator charged with an appropriate substance (in the case of ethanol, granular calcium chloride or silica gel). However, as Professor Black pointed out, aripiprazole is not a low melting point solid. 262 Vogel also teaches that drying can be accomplished using an electric oven controlled at a suitable temperature. In this connection, Vogel says: Many students place carefully recrystallised samples into a heated oven maintained at a temperature higher than the melting point of their solid with inevitable results; this leads to undue waste of effort and chemicals. Even if the melting point is known, it is always advisable to make a trial with a small quantity on a watch glass. In fact, a temperature of about 50°C over a period of 1-2 hours is usually adequate for the removal of the common organic solvents mentioned in Table 2.8. If the material can be left overnight at this temperature complete removal of water will occur. 263 Table 2.8 in Vogel identifies ethanol as one of the common organic solvents used for recrystallisation. 264 However, Vogel also acknowledges that, frequently, the solvent is so firmly held that it cannot be completely removed in a vacuum dessicator at ordinary temperatures and must be dried in a vacuum oven at higher temperatures. Vogel makes clear that the conditions for drying recrystallised material depend on the quantity of product, the nature of the solvent to be removed, and the sensitivity of the product to heat and to the atmosphere. 265 In criticising Associate Professor McGeary's further drying step, Professor Easton argued, secondly, that there was no reason to dry the crystals by heating above the boiling point of ethanol. He argued that, because of ethanol's vapour pressure, it is not necessary to boil it in order to evaporate it. He argued that there was simply no justification for Associate Professor McGeary drying his crystals at a temperature significantly above the boiling point of ethanol. 266 In response, Associate Professor McGeary acknowledged Vogel's teachings, but said that they provided only general guidance. He said that Vogel does not attempt to prescriptively set drying conditions; it recognises that a variety of conditions may be appropriate. I agree with Associate Professor McGeary's observations in respect of Vogel's teachings. 267 With respect to his selected range, Associate Professor McGeary said that it could be expected that heating at higher temperatures would remove solvent more quickly than heating at lower temperatures and that, in his experience, it was standard practice to heat recrystallised material above the boiling point of the solvent used in order to dry that material. He also noted that, as the melting point of aripiprazole obtained from following the directions in lines 5 to 15 of Example 1 was reported at around 139°C, he had no reason to suspect that drying the material within the range of 80°C to 100°C would decompose it. Finally, he said that he commenced his further drying step at 92°C, after setting the oven to 90°C, because 90°C was the midpoint in the range he had selected for step 4.9 in his protocol. 268 When challenged in cross-examination on why he selected this midpoint, Associate Professor McGeary made clear that he considered that heating anywhere in the range he had given in step 4.9 of his protocol would be appropriate to remove ethanol from the crystals. He gave this evidence: Well, let's assume you thought it was 80 degrees. Why didn't you just set the temperature at 80 degrees?---Well, there's an issue of - the temperature can affect how fast the drying occurs. So in general, the higher the temperature, the faster it comes off. But then there's also this issue of not wanting to overheat it for no particularly good reason. So the range of 80 to 100 is, as I've said, appropriate. Anywhere in that range would be acceptable to me. Why didn't you - what was the concern about overheating it?---I think it's just in general, it's best to do things at milder temperatures. Right - and what's the concern about overheating?---That you cook something unnecessarily, it may decompose. So you have that concern, do you?---Everyone has that concern. That's well known. So milder was better, in your view?---In general, but to be balanced against the rate -how fast does it happen. Well, if milder was better, what's wrong with 80 degrees?---Nothing's wrong with 80 degrees. Anywhere in that range, I thought, was appropriate. 80 degrees would have been appropriate. Well, if milder was better, in your view, and that was the view you had at the time, why did you select 90?---Because I thought there would be no harm to be done anywhere in the range of 80 to 100. If milder is better, why did you select 90?---Well, again, one weighs up rates, mildness, and whether anywhere in that range is appropriate. As I've said many times, anywhere in the range 80 to 100 is appropriate. I considered it appropriate then. But you wanted it to be quicker, did you?---Not necessarily. Well, what was the advantage of going to 90 over 80?---No particular advantage, as I've said. Anywhere in that temperature range would be okay. Well, I think you've indicated an answer to my questions. I won't ask you again, but when I asked you, "Why not 80," you said you didn't think 90 was going to do it any harm. Accepting that for present purposes, what was the benefit of 90, then - leaving aside whether you thought it was going to do any harm, what was the benefit of 90?---There was no particular benefit. 269 Professor Black's evidence provided general support for Associate Professor McGeary's evidence. Professor Black's evidence confirmed that drying in an oven at between 80°C and 100°C would have been an appropriate step to remove ethanol from the crystals. He said that, in his experience, if the material is being heated to remove solvent, the temperature used will be selected based on the properties, particularly the melting point, of the compound being recovered. He said that there was no reason to expect that, with a reported melting point of 139.0°C to 139.5°C, aripiprazole would be sensitive to heat in the region of 100°C. He expressed the general view that the oven drying carried out by Associate Professor McGeary was "a standard approach to removing solvent commonly used in September 2001". 270 Professor Black also gave evidence with respect to the McGeary experiment that, when Associate Professor McGeary undertook a melting point analysis of the crystals he had obtained, there was an observed change in the crystals in the region of 80°C to 100°C. Professor Black said that this change was consistent with the material being dried of ethanol. 271 Importantly, however, when discussing the White experiment in his second affidavit, Professor Black said that, if he had been carrying out the experiment, he would have left the material to dry in the vacuum oven overnight. He said that he would also have considered increasing the temperature to 50°C or 60°C. He said that, if the material was left in the oven overnight, he could "confidently conclude" that all the ethanol that could be removed by drying in the vacuum oven would have been removed. 272 Professor Black was cross-examined on this evidence. He confirmed that he regarded his own proposal as a reasonable approach. His evidence in cross-examination was that he would have left the material overnight in the vacuum oven at about 40°C. He continued: So I would regard that if [Associate Professor White] had have left it overnight at that temperature in the vacuum oven he would have got rid of all the ethanol. … There's not a lot of ethanol there. 273 Dr Rowe gave evidence that ethanol is a volatile compound, and not one that he would expect to be difficult to remove by common drying methods used in the manufacture of pharmaceuticals. In his second affidavit, he said: 13. … ethanol is a volatile compound and it is not a compound that I expect to be difficult to remove using common drying methods used in the manufacture of pharmaceuticals. In my experience, the most common drying techniques used in the manufacture of pharmaceuticals are tray drying (ie, spreading the compound on a tray in a conventional oven) or vacuum drying (drying the compound in a vacuum oven). The temperature used to dry the material will depend on the stability of the relevant pharmaceutical compound. In my experience the temperature is determined empirically. For example, the temperature may be determined using thermogravimetric analysis. If a sample of the material to be dried is analysed by thermogravimetric analysis the results obtained will indicate the temperature at which solvent is lost, and the temperature at which decomposition occurs … Although the actual temperature used will vary depending on the compound … in my experience common drying temperatures are around 50°C in a vacuum oven to remove surface adsorbed water or solvent, 80-90°C in a vacuum oven to remove a solvating compound or 80-85°C when tray drying in an oven. I would expect to be able to remove ethanol from aripiprazole using these techniques. 274 The evidence of the expert witnesses reveals a broad range of drying options by which, in their estimation, ethanol could be removed from the crystals obtained from following the directions in lines 5 to 15 of Example 1. This evidence is indicative of the range of options that would have been available before 25 September 2001 to the person skilled in the art seeking to obtain anhydrous aripiprazole crystals in their free base form when using the directions in lines 5 to 15 of Example 1 of the 141 application. 275 However, it is important to distinguish between (a) taking steps to remove ethanol from the crystals obtained by following the directions in lines 5 to 15 of Example 1 of the 141 application to obtain anhydrous aripiprazole crystals in their free base form, and (b) obtaining crystals having low hygroscopicity, as claimed in claim 12 of the patent. The McGeary experiment establishes that, by following the directions in lines 5 to 15 of Example 1, crystals having the essential characteristics of Crystals B and, in particular, the characteristic low hygroscopicity of Crystals B can be obtained if further drying is carried out for 24 hours in a vacuum oven at between 92°C and 94°C. But there is no evidence that shows that, by undertaking a different drying step selected from the range of options that would have been available to the person skilled in the art before 25 September 2001, crystals having that low hygroscopicity could be obtained or, if undertaken before the priority date, would have been obtained. Specifically, there is no evidence that, in relation to the crystals obtained by following the directions in lines 5 to 15 of Example 1: drying the crystals at 40°C in a vacuum oven overnight (the conditions under which Professor Black said he could confidently conclude that all the ethanol would be removed); or drying the crystals at 80°C to 90°C in a vacuum oven or at 80°C to 85°C on a drying tray in an oven (the conditions under which Dr Rowe said he would expect all the ethanol to be removed); or drying the crystals at any temperature within the range 80°C to 92°C in a vacuum oven for 24 hours (the conditions which Associate Professor McGeary said would also be appropriate to remove ethanol from the crystals), will result or would have resulted in crystals having the required low hygroscopicity. 276 As I have noted, the complete specification of the patent teaches that Crystals B will only be obtained by suitable heating. Reference Example 1 demonstrates that anhydrous aripiprazole crystals with a melting point of 140°C (matching the melting point of the anhydrous aripiprazole crystals obtained from the method in the Japanese unexamined patent publication corresponding to lines 5 to 15 of Example 1) exhibited unacceptable hygroscopicity of 3.28% after being dried for 40 hours at 80°C. There is, therefore, some evidence to suggest that, if the person skilled in the art were to undertake the drying step of Reference Example 1 to obtain aripiprazole in its free base form, he or she would not obtain Crystals B. On the other hand, the complete specification teaches that Crystals B can be obtained by heating conventional anhydrous aripiprazole crystals at 90°C to 125°C. Once again, reference can be made to Examples 5 and 6, where heating at 100°C and 120°C, respectively, was undertaken and yielded crystals having the characteristic low hygroscopicity of Crystals B. 277 In these circumstances, I am not persuaded that, by following the directions of lines 5 to 15 of Example 1 of the 141 application prior to 25 September 2001, the person skilled in the art would inevitably have obtained Crystals B when seeking to obtain anhydrous aripiprazole crystals in their free base form. In the course of following those directions, the person skilled in the art might have obtained Crystals B serendipitously, but it is equally possible that he or she would have obtained conventional anhydrous aripiprazole crystals with unacceptably high hygroscopicity. The 141 application does not teach, recommend or suggest, or otherwise disclose, any particular drying step. Still less is there any disclosure that (a) aripiprazole crystals can exhibit unacceptably high hygroscopicity, or (b) suitable heating can or should be undertaken to avoid obtaining crystals with unacceptably high hygroscopicity. In the words of General Tire, neither the directions in lines 5 to 15 of Example 1, nor more generally the 141 application, contain clear and unmistakeable directions to do what Otsuka claims to have invented. 278 For these reasons, I am not satisfied that the disclosure of the directions in lines 5 to 15 of Example 1 of the 141 application anticipates claim 12 of the patent. It follows that the other challenged claims are not anticipated by that prior art information. 279 I should also record that there is no evidence of the mean particle size of the crystals obtained when carrying out those directions. For this additional reason, I am not satisfied that the disclosure of the directions in lines 5 to 15 of Example 1 of the 141 application anticipates claim 16 of the patent.

[31]

The Aoki article and the Aoki poster 280 An initial question arises as to whether the Aoki article and the Aoki poster should be considered as a single source of information for the purpose of considering whether claims 12, 13, 14, and 16 of the patent are novel. 281 I am satisfied that they should be considered as a single source of information. Mr Aoki's evidence was that participants at the Symposium, which was held from 6 to 8 October 1996, were provided with a booklet that contained copies of papers, posters, and "other summaries" presented at the Symposium. Although a copy of the booklet is not in evidence, I am satisfied, on the basis of Mr Aoki's evidence, that it is more likely than not that both the Aoki article and the Aoki poster were included in that booklet. Moreover, Mr Aoki gave evidence to the effect that, at the Symposium, he stood with the poster displayed behind him, answering participants' questions. Given the circumstances in which the Aoki article and the Aoki poster were published at the Symposium, and given their content, I am satisfied that the person skilled in the art would have treated the Aoki article and the Aoki poster as a single source of information about the work that had been undertaken by Mr Aoki in analysing the characteristics of crystalline forms of aripiprazole. Certainly by each referring contemporaneously to, among other things, type 1 crystals and type 2 crystals of aripiprazole, the contents of the Aoki article and the Aoki poster show that they relate to that subject matter. Notwithstanding this conclusion, I should discuss what is relevantly disclosed by each publication. 282 The Aoki article discloses the existence of type 1, type 2, and type 3 crystals of aripiprazole. Type 1 and type 2 crystals are anhydrous. Type 3 crystals are hydrous. An XRPD pattern for each crystal type is disclosed in Figure 3 of the Aoki article. The XRPD pattern in Figure 3a is for type 1 crystals. Figure 3a contains only the shape of each pattern. 283 Dr Rowe's evidence was that the XRPD pattern disclosed in Figure 3a of the Aoki article for type 1 crystals is the same shape as the XRPD pattern for Crystals B given in Figure 5 of the complete specification of the patent. Dr Rowe also noted that the Aoki article discloses the melting point of type 1 crystals as 140°C. Crystals B exhibit an endothermic peak near about 141.5°C in TGA/DTA and an endothermic peak near about 140.7°C in DSC. Dr Rowe was of the opinion, based on his comparison of these data, that the type 1 crystals in the Aoki article and Crystals B are "the same crystalline form of aripiprazole". 284 Professor Withers gave evidence that the XRPD pattern for type 1 crystals disclosed in the Aoki article appeared to have similar characteristic peaks to the XRPD pattern given in Figure 5 of the complete specification for Crystals B. However, he said that, in the absence of a readable °2θ scale or, at the very least, an indicator of where the 10° and 20° °2θ positions are, as well as knowledge of the x-ray wavelength used, it was impossible to say whether the pattern in Figure 3a of the Aoki article and the pattern in Figure 5 of the complete specification are substantially the same. 285 Professor Withers continued: 78. Even though the relative heights of the peaks in the Fig. 3a [p]attern and the Figure 5 pattern appear to look similar, the relative heights of those peaks carries less weight than the position of the peaks along the °2θ axis in considering whether the XRPDs have the same crystalline form, for the reasons set out … above. 79. As far as I am aware there is no machine or software that could analyse the Fig. 3a [p]attern to determine whether it is the same as Figure 5 without the information listed … above. Without the °2θ scale, the Fig. 3a Pattern does not provide any information about the crystalline form of the sample being measured. 286 I prefer Professor Withers' evidence. It is based on greater expertise and more detailed reasoning. Dr Rowe's evidence in this regard appears to have been based on the assumption that a correspondence in shape between XRPD patterns necessarily includes a correspondence in value for the peaks shown in the patterns. Professor Withers' evidence makes clear that that is not necessarily so. I am not persuaded, therefore, that, for the purpose of assessing novelty, the Aoki article discloses with sufficient clarity that type 1 crystals have the same crystalline form as Crystals B. For this reason alone, the Aoki article cannot anticipate the invention claimed in claim 12 of the patent. 287 The Aoki poster provides DSC data and an XRPD pattern for type 1 crystals. There is no °2θ scale shown for the XRPD pattern. However, the five strongest peaks have been labelled with their °2θ positions. Although the values of these peaks differ slightly from those specified in the figures relating to claim 12 with respect to Crystals B, Dr Rowe considered these differences to be within standard experimental variation. Dr Rowe also noted that the Aoki poster disclosed an endothermic peak for type 1 crystals at 140°C measured by DSC. In his view, this was consistent with the endothermic peak of 140.7°C specified in claim 12. Dr Rowe was of the opinion, therefore, that type 1 crystals disclosed in the Aoki poster had the same crystalline form as Crystals B. 288 Professor Withers came to the same conclusion. In that connection, Professor Withers also noted that the relative peak heights in the XRPD pattern of type 1 crystals in the Aoki poster and in Figure 5 of the complete specification of the patent were very similar. 289 I am satisfied on this evidence that the Aoki poster discloses type 1 crystals that have the same crystalline form as Crystals B claimed in claim 12 of the patent. 290 As I have noted above in [89], Apotex's case was that a given crystalline form of aripiprazole will have, inevitably, a given hygroscopicity so that the disclosure of a particular crystalline form of aripiprazole will also inevitably disclose the compound's hygroscopicity in that form. Apotex argued that the disclosure of the XRPD pattern of type 1 crystals in the Aoki article or the Aoki poster, considered separately or in combination, also discloses Crystals B as claimed in claim 12 of the patent, including the characteristic low hygroscopicity of those crystals. It followed, on its argument, that claim 12 had been anticipated. 291 Apotex sought to develop this submission by the following argument: There is no novelty in merely testing and reporting the properties of a known substance. See, for example, Evans Medical Ltd's Patent. This follows as a matter of principle because what is claimed - a product - is not novel in the light of the published information. It does not matter if one of its inherent properties had not been expressly stated if the product itself had been disclosed. The absence of NMR, IR and hygroscopicity data does not change the fact that the same crystal form of aripiprazole as that claimed in the patent as Crystals B is clearly disclosed in each of the Aoki article and poster. 292 I do not accept Apotex's contention that it is sufficient, for the Aoki article or the Aoki poster to be novelty-destroying, that they disclose, separately or in combination, a particular crystalline form of aripiprazole whose physicochemical characteristics can simply be established. As is clear from my reasoning and findings in [106] to [109] above, I do not accept that a given crystalline form of aripiprazole will have, inevitably, a given hygroscopicity or that the disclosure of a given crystalline form of aripiprazole will inevitably disclose the compound's hygroscopicity in that form. 293 Although I am satisfied that the Aoki poster, and the Aoki article in combination with the Aoki poster, disclose type 1 crystals having the same crystalline form as Crystals B, the disclosures in those documents, nevertheless, do not anticipate claim 12 of the patent because they do not clearly disclose aripiprazole having the low hygroscopicity that is a characteristic of Crystals B. Whether a particular crystalline form of aripiprazole will possess the low hygroscopicity that is a characteristic of Crystals B will depend on whether the crystals have undergone certain processing steps, including suitable heating. The Aoki article and the Aoki poster are silent on whether the type 1 crystals there disclosed have undergone such heating. Moreover, as with the 141 application, the Aoki article and the Aoki poster do not disclose that (a) aripiprazole crystals can exhibit unacceptably high hygroscopicity, or (b) suitable heating should be undertaken to avoid obtaining crystals with unacceptably high hygroscopicity. 294 Apotex advanced an additional submission. It argued that the weight of the unchallenged evidence is that the chemistry involved in the synthesis of aripiprazole is straightforward. Apotex pointed to Associate Professor McGeary's evidence that, in the absence of a reported method for making a compound, he had routinely used the principles of retrosynthetic analysis - taught to third-year undergraduate students - to develop a synthetic pathway to the compound. Apotex argued that the Aoki article and the Aoki poster disclose not only the crystalline form of type 1 crystals but also that these crystals are obtained by recrystallisation from ethanol. It submitted that, armed with that information, the person skilled in the art would be in a position to undertake a recrystallisation of aripiprazole from ethanol. Apotex submitted that it should be inferred that, if that step was undertaken, the person skilled in the art would obtain crystals having the characteristics of Crystals B as claimed in claim 12 of the patent, just as Associate Professor McGeary had obtained such crystals when carrying out his experiment. 295 It is not necessary for me to deal with all aspects of that submission. It is sufficient to note that Apotex's argument must fail because there is nothing in the Aoki article or the Aoki poster that would instruct the person skilled in the art to carry out the specific drying step carried out by Associate Professor McGeary or any other drying step that would inevitably lead to crystals having the characteristic low hygroscopicity of Crystals B as claimed in claim 12 of the patent. 296 Additionally, the Aoki article and the Aoki poster do not disclose the mean particle size of type 1 crystals. It is accepted that particle size is not an inherent property of aripiprazole in a given crystalline form. It follows that, for this additional reason, the Aoki article and the Aoki poster, considered separately or in combination, do not anticipate claim 16 of the patent.

[32]

VALIDITY: FALSE SUGGESTION 297 A patent may be revoked on the ground that it was obtained by fraud, false suggestion or misrepresentation: s 138(3)(d) of the Act. Apotex did not allege fraud in the present case. It did, however, rely on false suggestion or misrepresentation. 298 The authorities make clear that the alleged false suggestion or misrepresentation may be contained in statements made in the complete specification or reside in conduct occurring in the course of applying for the patent. It is enough that the statement or conduct is a "material inducing factor" leading to the grant of the patent. The relevant principles are discussed in Prestige Group (Australia) Pty Ltd v Dart Industries Inc (1990) 26 FCR 197 at 198-201 and 213-218. 299 Here, Apotex relied on statements in the complete specification and statements made to the Commissioner of Patents (the Commissioner) by Otsuka's patent attorneys in the course of prosecuting the application for the patent. It provided particulars of this allegation in its particulars of invalidity dated 27 November 2009. These particulars appeared unchanged in its amended particulars of invalidity dated 7 September 2010 and its further amended particulars of invalidity dated 27 June 2011. In these particulars, Apotex referred to certain passages in the complete specification and correspondence from Otsuka's attorneys to the Commissioner which it said amounted in effect to representations that the prior art forms of aripiprazole were undesirably hygroscopic for the purposes of pharmaceutical formulation. It contended that these representations - which, it argued, were material inducing factors in obtaining the patent - were false because "prior art forms of aripiprazole" were not undesirably hygroscopic. 300 In its particulars, Apotex identified the following statements in the complete specification as constituting part of the false suggestion or misrepresentation: … that anhydrous aripiprazole crystals obtained by the aforementioned [prior art] methods have the disadvantage of being significantly hygroscopic. The hygroscopicity of these crystals makes them difficult to handle since costly and burdensome measures must be taken in order to ensure they are not exposed to moisture during process and formulation. Exposed to moisture, the anhydrous form can take on water and convert to a hydrous form. This presents several disadvantages… It would be highly desirable to discover a form of aripiprazole that possessed low hygroscopicity thereby facilitating pharmaceutical processing and formulation operations required for producing dosage units of an aripiprazole medicinal product having improved shelf-life, suitable dissolubility and suitable bioavailability. 301 It also identified statements made in correspondence from Otsuka's patent attorneys to the Commissioner on 12 February 2004 (the February correspondence) and 6 August 2004 (the August correspondence) as constituting part of the false suggestion or misrepresentation. 302 The relevant statements in the February correspondence were: … the aripiprazole prepared by the process described in the prior references … has a problem of "high hygroscopicity", and there is a serious problem that the known aripiprazole is easily converted to the hydrate, even in a pharmaceutical preparation. The aripiprazole prepared by the process described … has a high hygroscopicity as well as the above-mentioned problem. The elusion property of the aripiprazole hydrates thus formed is quite inferior to that of the aripiprazole anhydrates crystalline forms, as is demonstrated in page 98, Table 2 of the present specification. Under the circumstances, the present inventors have extensively conducted research in order to solve the above-mentioned problem and obtain a low hygroscopic aripiprazole which does not substantially convert to the hydrates… None of the cited references teach the aripiprazole with low hygroscopicity of the present invention and process for the preparation thereof. The aripiprazole with low hygroscopicity was not able to be obtained before the present invention, so that the above-mentioned problem may be solved by the present invention for the first time. Further, the novel low hygroscopic aripiprazole according to the present invention has significant industrial applicability. … D3 does not teach or suggest the above-mentioned process described in the present specification. 303 The relevant statements in the August correspondence were: … the Aripiprazole Hydrate A of the present invention is clearly distinguishable from the conventional aripiprazole hydrate… The cited references D1 to D3 are quite silent as to the present Aripiprazole Hydrate A… Further, the present Aripiprazole Hydrate A is useful as an intermediate in the preparation of the anhydrous crystalline forms having low hygroscopicity, i.e., Anhydrous Aripiprazole Crystals B of the present invention. Of course, D1 to D3 neither disclose nor suggest the low hygroscopicity crystals or that Hydrate A is an intermediate in their preparation. … … Anhydrous Aripiprazole Crystals of the present invention … have extremely low hygroscopicity, as compared to the conventional anhydrous aripiprazole crystals … That is, the conventional anhydrous aripiprazole crystals have "high hygroscopicity" (see Table 1), so that it is easily converted into the hydrate due to the "high hygroscopicity" even in a pharmaceutical preparation. Further, the dissolution property of the aripiprazole hydrates thus formed is quite inferior to that of the anhydrous aripiprazole crystalline forms according to the present invention … 304 In the February correspondence and the August correspondence, the reference to D2 is the 141 application and the reference to D3 is the Aoki article. The reference to D1 is to a publication that does not feature in the present case. 305 On 19 November 2012, Apotex was ordered to file and serve a document whose purpose was to give notice of the way it sought to put its case on false suggestion. This order was made to require Apotex to refine the particulars it had previously given. The substance of Apotex's notice, provided on 21 November 2012, was that: The "form of aripiprazole", the "low hygroscopic aripiprazole", and the "Anhydrous Aripiprazole Crystals B" referred to in the statements quoted above is a crystalline form of aripiprazole referred to in the complete specification as "Anhydrous Aripiprazole Crystals B". The prior art forms of aripiprazole referred to by Apotex in its particulars were the products referred to in lines 5 to 15 of Example 1 of the 141 application, and the type 1 crystals referred to in the Aoki article and the Aoki poster. Each of these prior art forms is the polymorphic form of aripiprazole identified by the XRPD spectrum in Figure 5 of the complete specification of the patent and/or possesses the characteristic peaks referred to in line 9 of page 24 of the complete specification. Each of these prior art forms is the same polymorph of aripiprazole as Anhydrous Aripiprazole Crystals B or possesses the same crystalline form. These prior art forms are not undesirably hygroscopic. 306 The essence of Apotex's contention on this aspect of its case is captured in the following submission: Although "Type-I" is the very same polymorph as "Crystals B", the patent specification re-badges it with the result that it represents or suggests a difference from the prior art. In truth, that difference does not exist. 307 It can be seen that this aspect of Apotex's challenge to the patent depends on the acceptance of its contentions that the crystals obtained from following the directions in lines 5 to 15 of Example 1 of the 141 application have the same crystalline form as type 1 crystals referred to in the Aoki article and the Aoki poster and that, in turn, type 1 crystals have the same crystalline form that is one of the characteristics of Crystals B claimed in claim 12 of the patent. Importantly, it also depends on the acceptance of Apotex's contention that a given crystalline form of aripiprazole will have, inevitably, a given hygroscopicity. 308 As recorded above, I accept that type 1 anhydrous aripiprazole crystals and Crystals B have the same crystalline form. Accordingly, I accept the first part of Apotex's argument. However, I do not accept that a given crystalline form of aripiprazole will have, inevitably, a given hygroscopicity. It follows that Apotex's challenge based on false suggestion or misrepresentation cannot succeed. For completeness, I should record that the statements relied on by Apotex do not represent or suggest that Otsuka had discovered a new crystalline form of aripiprazole, only that it had solved the problem of hygroscopicity. 309 There was, however, an additional aspect to Apotex's challenge to which I should refer. In final submissions, Apotex argued that, according to one experiment disclosed in Otsuka's internal experimental reports discovered in this proceeding, a sample of aripiprazole with low hygroscopicity had been obtained when crystals were dried at 80°C. Apotex then pointed to the following statement in the February correspondence: … The aripiprazole with low hygroscopicity was not able to be obtained before the present invention, so that the above-mentioned problem may be solved by the present invention for the first time. … 310 Apotex submitted that this statement must be a false suggestion or misrepresentation in light of the results of the earlier experiment it had identified. 311 The first matter to be noted is that Apotex did not signify, prior to the hearing, that it relied on the results of this experiment for the purpose of propounding its challenge to the patent on the basis of false suggestion or misrepresentation. Its case was confined by its particulars as refined by the note that it was ordered to provide. The sample from the experiment on which Apotex sought to rely was not one of the prior art forms it had identified. Moreover, this sample was not produced in accordance with the directions of lines 5 to 15 of Example 1 or by the method briefly reported in the Aoki article and the Aoki poster. I do not propose to allow Apotex to now rely on this experiment to argue that the patent was obtained on a false suggestion or misrepresentation. It is plainly additional matter for which proper notice has not been given. 312 However, even if Apotex were allowed to advance its case on false suggestion or misrepresentation by reference to the results of that particular experiment, its challenge could not succeed in any event, for the following reasons. 313 The February correspondence was a reply to the Examiner's First Report dated 13 August 2003. Relevantly for present purposes, that report raised the objection that the 141 application disclosed aripiprazole and thus anticipated all the filed claims. It also raised the objection that two claims directed to a process for the preparation of granules and two claims directed to a process for making a pharmaceutical solid oral dosage preparation were anticipated by the Aoki article. 314 The quoted passage from the February correspondence set out in [309] above was addressed to the objection based on the 141 application, namely, that aripiprazole had previously been disclosed. When that particular passage in the February correspondence refers to the invention solving the problem of hygroscopicity "for the first time", the author is plainly referring to the fact that the 141 application does not address the problem of hygroscopicity. In my view, the impugned statement is perfectly correct and contains no false suggestion or misrepresentation. The same may be said of the reference in the February correspondence to the Aoki article. That article does not address the problem of hygroscopicity. The other publication cited by the examiner is not in evidence. It has not been suggested that it is relevant to the present question. 315 For present purposes, it is not to the point that, when conducting its own internal research before applying for the patent, Otsuka may have found a low hygroscopic form of aripiprazole. The point made by the February correspondence is that the invention of the low hygroscopic form of aripiprazole was novel because it had not previously been made publicly available by the publications cited in the Examiner's First Report. That point is not gainsaid by the results of the experiment on which Apotex would now seek to rely. 316 There is, however, an additional matter to note. In a subsequent experiment conducted by Otsuka in the period of April to May 2001, a sample of anhydrous aripiprazole crystals produced by the same process (namely, by pulverising aripiprazole hydrate and then drying the pulverised product at 80°C) exhibited unacceptable hygroscopicity, outside the terms of claim 12. This casts doubt on the reliability of the results of the humidity test with respect to the sample obtained from the earlier experiment. It can be seen, therefore, that the results of the particular experiment on which Apotex would now seek to rely are contentious. This underscores the reason why Apotex should not be permitted to rely on this particular experiment now, when, through lack of proper notice, the applicants have not been afforded a full opportunity to deal with it in evidence.

[33]

Validity: inventive step 317 An invention is only a patentable invention for the purposes of a standard patent if the invention, as claimed, involves an inventive step when compared with the prior art base as it existed before the priority date of the claim: s 18(1)(b)(ii) of the Act. 318 Subsections 7(2) and 7(3) of the Act are important in determining when an invention will be taken to involve an inventive step. 319 Subsection 7(2), as applicable to the present case, is as follows: For the purposes of this Act, an invention is to be taken to involve an inventive step when compared with the prior art base unless the invention would have been obvious to a person skilled in the relevant art in the light of the common general knowledge as it existed in the patent area before the priority date of the relevant claim, whether that knowledge is considered separately or together with the information mentioned in subsection (3). 320 Subsection 7(3), as applicable to the present case, is as follows: The information for the purposes of subsection (2) is: (a) any single piece of prior art information; or (b) a combination of any 2 or more pieces of prior art information; being information that the skilled person mentioned in subsection (2) could, before the priority date of the relevant claim, be reasonably expected to have ascertained, understood, regarded as relevant and, in the case of information mentioned in paragraph (b), combined as mentioned in that paragraph. 321 Apotex contended that the invention claimed in each of the challenged claims does not involve an inventive step. It contended that the question should be approached by adopting a problem-solution approach, where the solution is the invention as claimed. It relied, in part, on certain observations made by Aickin J in The Wellcome Foundation Limited v V.R. Laboratories (Aust.) Proprietary Limited (1981) 148 CLR 262. In that case, the High Court was dealing with the question of whether discovery should be ordered of documents relating to the patentee's research and experiments leading to the invention as claimed. This involved consideration of the admissibility of evidence of a patentee's own research and experimentation for the purpose of demonstrating that, at the relevant time, the invention was obvious. His Honour said (at 286): … In more recent authorities discovery has been ordered to enable those attacking the patent to search for material which may suggest that all that the inventor actually did was to take a series of routine steps or make a series of routine experiments. It is still correct to say that a valid patent may be obtained for something stumbled upon by accident, remembered from a dream or imported from abroad, if it otherwise satisfies the requirements of the legislation. What is important is that the patent itself should involve an inventive step, whether or not it was consciously taken by the patentee and whether or not it appeared obvious to the patentee himself. … 322 His Honour followed this by stating: … The test is whether the hypothetical addressee faced with the same problem would have taken as a matter of routine whatever steps might have led from the prior art to the invention, whether they be the steps of the inventor or not. [Emphasis added] 323 The applicability, in appropriate cases, of a problem-solution approach to determining whether a claimed invention involves an inventive step is supported by a number of authorities. These authorities are discussed in Apotex Pty Ltd (ACN 096 916 148) v Sanofi-Aventis and Others (2009) 82 IPR 416 at [145]-[159]. It is not necessary to repeat that discussion for present purposes save to note that, as Bennett and Middleton JJ cautioned (at [159]), the analysis of whether an invention involves an inventive step is not confined to a problem-solution approach. 324 Apotex accepted that a problem-solution approach might not always be appropriate. It argued, however, that such an approach was appropriate in the present case because a problem is stated on the face of the complete specification: "prior art" aripiprazole is undesirably hygroscopic. 325 Apotex then advanced what it identified as a separate argument: s 18(1)(b)(ii) of the Act effected a change in the law with respect to identifying an inventive step by invoking a comparison between the invention as claimed and the prior art base, which relevantly includes a document that is publicly available, whether in or out of the patent area, and whether or not the information in it is part of the common general knowledge. In developing this argument, Apotex pointed to the same comparison expressed in the opening words of s 7(2) of the Act: "… an invention is to be taken to involve an inventive step when compared with the prior art base …". Apotex submitted that there is now a statutory starting point (the prior art base) and a finishing point (the invention as claimed) which must be compared to identify the step to be assessed for inventiveness in light of the common general knowledge. 326 It is not clear to me how this separate argument sits with the problem-solution calculus of its first argument, which posits the problem and not the prior art base as the starting point for identifying and assessing the step whose inventiveness is in issue. It is not necessary, however, for me to dwell on whether and, if so, how the two arguments can or should be reconciled because I am not persuaded that Apotex's separate argument should be accepted. It has been rejected on at least three prior occasions in first instance decisions of the Court: Sanofi-Aventis Australia Pty Ltd and Others v Apotex Pty Ltd (No 3) (2011) 196 FCR 1 at [229]-[230]; Wake Forest University Health Sciences and Others v Smith & Nephew Pty Ltd (ACN 000 087 507) and Others (No 2) (2011) 92 IPR 496 at [701]-[717]; Apotex Pty Ltd v AstraZeneca AB and Another (No 4) (2013) 100 IPR 285 at [216]-[217]. I am not persuaded that those decisions are plainly wrong. It is appropriate, therefore, that I follow them. 327 The rejection of Apotex's additional argument is, perhaps, of no real consequence in the present case when regard is had to the way in which Apotex developed its first argument. Apotex submitted: The 141 application forms part of the starting point. Its equivalent - the Japanese unexamined patent publication - is cited in the complete specification of the patent as background to the invention. The 141 application is, in any event, a document that would be ascertained, understood, and regarded as relevant by the person skilled in the art and would, therefore, be considered: s 7(3) of the Act. There is no inventive step in removing ethanol from the product of recrystallisation obtained by following the directions of lines 5 to 15 of Example 1 of the 141 application. 328 Apotex submitted that the present case is unusual because "there is no question, as a matter of fact, that the very polymorph that is claimed forms part of the problem, or the starting point: Mr Aoki's Type-I crystals". Apotex continued: … To the extent that an improved hygroscopicity profile is claimed, it was simply an inherent characteristic of that polymorph. Otsuka may have invented a method of preventing the conversion of type-I to the monohydrate by milling first then drying, but that does not confer novelty on Type-I renamed as Crystals B. Claims 12, 13 and 16 are not confined by questions of purity or industrial scale. … 329 Apotex also submitted: … There was no barrier to cross in the context of an inventive step analysis of the product, per se. 330 The applicants submitted that a problem-solution approach should not be adopted when determining, in the present case, whether the invention as claimed involves an inventive step. The applicants submitted that: … there are no words in the legislation which require that there be a problem to be solved before one can have an inventive step. The issue is at large. The question is whether it was obvious to move from the common general knowledge to the claimed invention. The common general knowledge may or may not include a "problem" to be solved. A recognised objective indicium of invention is the circumstance where a problem has been recognised in the common general knowledge as requiring a solution which is satisfied by the claimed invention. That is an example of "longfelt want". However, an objective indicium of invention can also be the satisfaction of an "unfelt want" as the High Court recognised in Wellcome Foundation Ltd v VR Laboratories (Aust) Pty Ltd (1981) 148 CLR 262 at 287.4. In such a case, part of the inventive step includes the perception that although not recognised by others in the field, a particular product or method could provide a key and distinct therapeutic advantage. [Footnote omitted] 331 With specific reference to the application of the deeming provision of s 7(2) of the Act, the applicants submitted: … the "starting point" approach is incompatible with the provisions of the Act and thus incorrect as a matter of principle. Inventive step is to be assessed in accordance with s 7(2). That provision deems an inventive step to be present "unless the invention would have been obvious to a person skilled in the relevant art" in the light of the common general knowledge with any available s 7(3) information. There is no room for any resort to information which is not either part of common general knowledge or available under s 7(3). In particular, there is no room for any resort to information simply because it is referred to in the specification of the patent, whether it be expressed as a "starting point" or problem to be addressed or otherwise. 332 The applicants also pointed to two factual matters. They submitted that Apotex had not proved that either aripiprazole or the hygroscopicity problem associated with conventional anhydrous aripiprazole crystals were part of the common general knowledge before 25 September 2001. I accept that these facts have not been established. 333 I do not find it necessary to resolve the parties' competing positions on the correct framework to be used when considering the question of inventive step in the present case. I will proceed by assuming, but without deciding, that a problem-solution approach is the correct framework to adopt in the present case, in order to evaluate the cogency of Apotex's case on obviousness advanced on that particular basis. Proceeding in that way, the question then arises as to what information is to be taken as comprising the starting point. Undoubtedly, knowledge of the existence of conventional anhydrous aripiprazole crystals and the fact that those crystals can exhibit unacceptably high hygroscopicity would be part of the starting point. However, I doubt that knowledge of the 141 application itself would be part of the starting point simply because, as Apotex contended, it is cited in the complete specification as background to the invention. 334 Nevertheless, I am prepared to find, perhaps somewhat generously in Apotex's favour, that the person skilled in the art would reasonably be expected to have ascertained the existence of the 141 application. Dr Rowe gave evidence that if, as a pharmaceutical formulation chemist, he was faced with the problem of overcoming the disadvantage that anhydrous aripiprazole crystals are significantly hygroscopic, he would review, as a first step, the physicochemical information relating to aripiprazole. If not provided with that information, he said that, in 2001, he would have requested a literature search to be undertaken on his behalf by a specialised searcher using search terms that he had specified. He referred to the existence of electronic databases that index journal articles and patents relating to pharmaceuticals. There is no evidence, however, of the likely results that would have been obtained by conducting a search of the electronic databases that Dr Rowe identified using appropriate search terms prior to 25 September 2001. 335 Professor Easton gave evidence of carrying out a literature search using SciFinder. SciFinder is an online database of chemistry and other science-related information, containing research papers, information about chemical substances, chemical suppliers, and the like. Professor Easton said that, if carrying out a literature search in September 2001, he would have used SciFinder. In the present case, he searched for references to aripiprazole, its molecular formula, and its CAS registry number through the "substance identifier" feature of SciFinder. A CAS registry number is an identifying number assigned to a compound when it is first included in the SciFinder database. The 141 application was one of the documents he found searching in this manner. 336 Although SciFinder is not a source to which Dr Rowe referred, I am persuaded by Dr Rowe's evidence and Professor Easton's evidence that electronic search tools were available to the person skilled in the art as at 25 September 2001 and that, in all likelihood, resources of that nature would have been used by the person skilled in the art seeking to obtain information with respect to aripiprazole. I am further persuaded by Professor Easton's evidence that, on balance, the person skilled in the art undertaking such a search at that time would have been likely to have found the 141 application. 337 I am also satisfied that if, prior to 25 September 2001, the person skilled in the art had located the 141 application, he or she would have understood it and regarded it to be relevant to aripiprazole. 338 Based on these findings, it does not matter whether, as Apotex contended, the 141 application is taken as part of the starting point. It is certainly prior art information that, by dint of s 7(3) of the Act, is to be taken to be available to the person skilled in the art for the purpose of applying s 7(2) of the Act. However, as I have found above, there is no disclosure in the 141 application that aripiprazole crystals can exhibit unacceptably high hygroscopicity or that suitable heating can or should be undertaken to avoid obtaining crystals with unacceptably high hygroscopicity. In my view, if, before 25 September 2001, the person skilled in the art were to be presented with the 141 application, he or she would be no wiser about the postulated problem of significant hygroscopicity in anhydrous aripiprazole crystals or how to solve that problem. 339 Apotex's submission proceeds on the assumption that the person skilled in the art, on being armed with the 141 application, would be motivated to "remove" ethanol. I have accepted that the person skilled in the art might be motivated to obtain anhydrous aripiprazole crystals in their free base form: see [256] to [257] above. But this would not be to address a problem of significant hygroscopicity. Perhaps more importantly, Apotex's submission proceeds on the basis that the removal of ethanol to obtain anhydrous aripiprazole crystals in their free base form is equivalent to achieving anhydrous aripiprazole crystals having the characteristic low hygroscopicity of Crystals B. I do not accept that to be the case. Once again, Apotex's case proceeds on its central contention, which I have rejected, that a given crystalline form of aripiprazole will have, inevitably, a given hygroscopicity. Moreover, as I have found, I am not persuaded that, by following the directions of lines 5 to 15 of Example 1 of the 141 application prior to 25 September 2001, the person skilled in the art would inevitably have obtained Crystals B when seeking to obtain anhydrous aripiprazole crystals in their free base form: see [277] above. In my view, therefore, the 141 application does not advance Apotex's case on lack of inventive step in any material way. 340 In my view, even adopting Apotex's preferred framework of a problem-solution approach, the invention, as relevantly claimed, is far from obvious. Using that framework, the question would be whether, before 25 September 2001, the person skilled in the art would be directly led as a matter of course to take the steps that result in Crystals B with their characteristic low hygroscopicity, in the expectation that those steps might well produce those crystals: Aktiebolaget Hässle and Another v Alphapharm Pty Limited (2002) 212 CLR 411 at [50]-[53]; Alphapharm Pty Ltd (ACN 002 359 739) v H Lundbeck A/S and Another (2008) 76 IPR 618 at [180]; Sigma Pharmaceuticals (Australia) Pty Ltd v Wyeth and Another (2010) 88 IPR 459 at [251]-[252]. 341 Apotex's case on lack of inventive step was advanced principally through Dr Rowe's evidence. I turn to consider that evidence. 342 Dr Rowe said that, in his experience, hygroscopicity issues were, in the work of pharmaceutical formulation chemists, addressed in the formulation process by chemical means or physical means, or a combination of both. 343 In his first affidavit, Dr Rowe gave the following evidence with respect to chemical means that would be adopted to address hygroscopicity issues: 53. Addressing hygroscopicity issues by chemical means involves identifying a form (ie, a polymorph, solvate or salt) of the pharmaceutical compound which is not hygroscopic or is less hygroscopic. Different polymorphs, solvates and salts of a pharmaceutical compound commonly have different physicochemical properties, such as different degrees of hygroscopicity … 54. If a non-hygroscopic form of the pharmaceutical compound was identified in preformulation studies, this form, whether it be a particular polymorph, solvate or salt, would normally have been used in the pharmaceutical formulation, subject to its overall suitability (for example, having suitable bioavailability). 55. In my experience as at September 2001, information about polymorphic forms of a pharmaceutical was routinely obtained during the preformulation stages of pharmaceutical development … This information was routinely obtained and available to a formulation chemist working with the pharmaceutical because it was required to develop a pharmaceutical and to satisfy regulatory requirements … 344 Later in his first affidavit, Dr Rowe said that, absent relevant information obtained by search, he would have obtained analytical studies from a sample of aripiprazole to determine its physicochemical properties. In that connection, he gave this evidence: 63. These analytical studies would include studying the existence of polymorphs of aripiprazole. As part of these analyses, I would have attempted to prepare different polymorphs of aripiprazole by: (a) Heating samples of aripiprazole to determine whether a polymorphic conversion occurred. (b) Recrystallising samples of aripiprazole from different solvents to determine whether different crystalline forms were recrystallised from different solvents. In September 2001, ethanol would have been the first solvent I would have used, and I would have used both pure ethanol, and aqueous ethanol (ie, ethanol containing water). Other solvents I would have used as at September 2001 after ethanol were acetonitrile, ethyl acetate, butyl acetate, chloroform, dichloromethane and hexane. (c) Studying the effect of grinding and compression on the polymorphic form of aripiprazole. 345 He also gave this evidence: 65. If the polymorphic studies indicated that aripiprazole was capable of existing in more than one polymorphic form, I would have studied the physicochemical properties, including hygroscopicity, of each polymorph. For example, to study hygroscopicity I would have exposed samples of different polymorphs of aripiprazole to various levels of moisture and temperature to determine how much moisture they take up, and which polymorphs are less hygroscopic and more suitable for development to address the Aripiprazole Scenario. If aripiprazole existed in more than one polymorphic form, I would have expected to identify different polymorphic forms of aripiprazole. I would also have expected that some of the polymorphs would be less hygroscopic than other polymorphs. 346 This evidence suggests that, even with a focus on polymorphs, it would have been necessary for the person skilled in the art, before 25 September 2001, to have engaged in a significant research project to attempt to discover how the hygroscopicity problem with anhydrous aripiprazole crystals might be solved using different polymorphic forms of aripiprazole. It does not suggest an obvious solution to a given problem. Indeed, it bespeaks no more than the speculative possibility of a solution which is presently unknown and, perhaps, unattainable. 347 Dr Rowe was cross-examined on this evidence insofar as it relates to an investigation of different polymorphs as a chemical means to overcome hygroscopicity issues. He was confronted with evidence given by him in another proceeding in which he said that, in his experience, polymorphs are not considered for their impact on hygroscopicity. To say the very least, that view sits uncomfortably - indeed, discordantly - with Dr Rowe's evidence quoted above. Under cross-examination, Dr Rowe effectively resiled from the thrust of his evidence-in-chief quoted above. He confirmed for the purposes of the present case that, as a general rule, polymorphs are not considered for their impact on hygroscopicity. Indeed, he said that, in his experience, different polymorphs which he had produced "have not shown great differences in hygroscopicity". In the course of seeking to explain why a discussion of polymorphs featured so prominently in his first affidavit, as quoted above, Dr Rowe accepted that his evidence was influenced by an assumption he made or inference he drew that the patent in suit was one concerning polymorphs of aripiprazole. He said that, but for that assumption or inference, the above quoted passages might "have read differently". This explanation does not, however, engage why Dr Rowe was prepared to advance, as matters of fact, seemingly contradictory positions with respect to the utility of considering different polymorphic forms of a given compound in order to overcome a problem of hygroscopicity exhibited by that compound. 348 It then emerged in cross-examination that, if confronted with the problem of hygroscopicity as at September 2001, Dr Rowe's first approach would have been to consider simple physical means, about which he gave the following evidence in his first affidavit: 56. Depending on the nature of the hygroscopicity issues, hygroscopicity was, in my experience as at September 2001, often controlled in manufacture and storage of a pharmaceutical composition. There were a number of techniques routinely used, if appropriate, to physically control hygroscopicity. 57. As at September 2001, there were, in my experience, a variety of physical measures available to control hygroscopicity. For example, the following measures were used: (a) Manufacturing the pharmaceutical composition in a low humidity environment, ie, an environment in which there is very little or no water vapour present in the air. The absence of water in the environment means that the hygroscopic pharmaceutical compound cannot take up water from the environment. (b) Selecting excipients to include in the formulation, such as microcrystalline cellulose and mannitol, that tend to prevent moisture being taken up by the pharmaceutical and the pharmaceutical formulation. (c) Coating the particles of the pharmaceutical to be formulated with a water impermeable coat (microencapsulation). (d) Coating the pharmaceutical composition (eg, a tablet) with a water impermeable substance to prevent water from permeating into the tablet. (e) Adding silica gel, for example in a sealed sachet, to the container in which the formulated product is stored. 58. These physical means can alleviate or minimise the hygroscopicity problem. However, they also increase the cost of production and storage (as relevant) of the pharmaceutical composition. In the case of some hygroscopic compounds these physical means cannot be used to control hygroscopicity to the extent necessary to prepare a solid oral dosage form. 349 He accepted that the above means would be one obvious solution to the problem of hygroscopicity. When asked in re-examination why his first approach would be to consider physical means, he said: … Well, they're, in fact, the simplest way and, certainly, in my experience and, also, I believe, my [colleagues'] experience, most of the problems can be solved by simple physical means. And by that I mean things like low humidity manufacturing areas, the use of oil packaging in the final tablets, the use of certain excipients and silicone dioxide to mop up moisture and so forth. And then once the drug in a dosage form is in a foil pack, the likelihood of water getting through is extremely small. But that would be borne out in stability studies. That's - that's purely from the point of view of ease rather than looking at something else because if you look at a new salt or polymorph, to the TGA that's a new drug substance, and you will have to do more viability studies, stability studies, it's very expensive and time consuming. 350 He accepted that another obvious solution would be to attempt to use a salt form of the compound. He said that this would have been his personal preference as at September 2001 after having first investigated physical means to solve the problem. Tellingly, Dr Rowe accepted that it would not have occurred to him, as a solution to the problem, to "stay within the same polymorph" and to heat it. 351 Professor Easton agreed that the physical means proposed by Dr Rowe could be possible means for controlling hygroscopicity with respect to aripiprazole. He said that, alternatively, he would have considered investigating a liquid formulation for the drug. He said that, if that approach was not successful, he would consider different forms of aripiprazole, being salts, polymorphs, and solvates, with the investigation of salts being his "strong preference". He said that he would not have given high priority to investigating solvates and polymorphs and that approach was not, in his experience, routine in September 2001 or now. 352 Nevertheless, he gave the following evidence in his second affidavit: 105. If I had ultimately decided to start investigating solvates and polymorphs as a way to address the problems identified in lines 1-3 of the [patent], I would have first tried to recrystallise from different solvents. That process is simply trial and error, and might or might not result in a different solvate or polymorph. 106. In my experience, recrystallisation is the most common way to investigate whether a new polymorph exists. I am also aware of trying, and have in the past tried, sublimation, pressure, heating and grinding. Sublimation is heating under reduced pressure so that a material converts from solid directly to vapour, then condenses directly to solid, so as to minimise the likelihood of decomposition associated with liquid intermediate phases. 107. I would have tried grinding because I know from my experience with cyclodextrins that it might work. However, from my experience, I would not have expected this approach to have a high likelihood of success because I regularly grind materials to obtain particular particle sizes and do not see polymorphs. … grinding is also used to prepare samples for XRPD analysis and the assumption is that the grinding will not change the crystal form. 108. I would have tried heating only as a last resort after trying recrystallisation from different solvents because heating generally leads only to melting and perhaps decomposition. If and when a solid reforms by condensing the melted liquid, it is common to obtain a solid block or amorphous material, rather than a free-flowing crystalline product. 353 Professor Easton accepted in cross-examination that, if melting and decomposition were concerns on heating, it was possible to test a small sample before proceeding. Notwithstanding that acceptance, Professor Easton's evidence strongly supports the conclusion that the solution to the unacceptable hygroscopicity of anhydrous aripiprazole crystals was far from obvious. 354 Finally, the evidence of Otsuka's own research effort assists in concluding that the solution was far from obvious. The applicants pointed to Mr Aoki's evidence, supported by Otsuka's internal research reports, that it took many years to produce a low hygroscopic form of aripiprazole. The applicants submitted that the "road to a low hygroscopic form of aripiprazole was long and arduous and involved multiple scientists, many years of experimentation, including multiple clinical studies, with failures and a range of problems along the way". I accept that submission. 355 In Hässle at [58], the High Court said: 58 … The tracing of a course of action which was complex and detailed, as well as laborious, with a good deal of trial and error, with dead ends and the retracing of steps is not the taking of routine steps to which the hypothetical formulator was taken as a matter of course. In [In the Matter of I.G. Farbenindustrie A.G.'s Patents [1930] 47] RPC 289 at 322, Maugham J had said that while "mere verification is not invention", what he likened to the citadel of invention "may be captured either by a brilliant coup-de-main or by a slow and laborious approach by sap and mine according to the rules of the art; the reward is the same." 356 For these reasons, Apotex's contention that the challenged claims do not involve an inventive step does not succeed. Absent the adoption of a problem-solution approach, it follows, even more so, that Apotex's contention cannot succeed. As I have noted, Apotex did not prove that either aripiprazole or the hygroscopicity problem associated with anhydrous aripiprazole crystals were part of the common general knowledge before 25 September 2001. 357 There is a remaining issue to which I should refer concerning the challenge to claim 16. It is to be recalled that claim 16 claims Crystals B with a mean particle size of 50 microns or less. Apotex contended that milling to a size of less than 50 microns was straightforward to the person skilled in the art as at 25 September 2001. Implicit in that contention was that milling to a size of less than 50 microns would not itself provide any missing inventive step. It is not necessary to deal with that contention because I am satisfied that the combination claimed by claim 16, including, as it does, Crystals B, does involve an inventive step. 358 A related argument advanced by Apotex was that there is no working interrelationship between Crystals B and particle size - the crystal form in question is simply made smaller by means well within the common general knowledge. Apotex contended that claim 16 "is not a truly new combination". I do not accept that argument. Dr Rowe's evidence, which Apotex advanced, was that the particle size of the pharmaceutical can influence the rate of dissolution in the body and can therefore impact on bioavailability. He said that particle size may also affect the manufacturing process, in particular, the flow properties of a powder during manufacture, the distribution of the pharmaceutical within a pharmaceutical formulation, and the stability of a pharmaceutical. This evidence removes the foundation for Apotex's contention that the combination claimed in claim 16 of the patent is not a proper combination.

[34]

Validity: manner of manufacture 359 An invention is only a patentable invention for the purposes of a standard patent if the invention, as claimed, is a manner of manufacture within the meaning of s 6 of the Statute of Monopolies: s 18(1)(a) of the Act. 360 Apotex originally directed this challenge to validity to a large number of claims. However, in the course of submissions, the challenge came to be limited to claims 35, 36, 43, 44, and 45. So confined, this challenge to validity has two aspects. 361 The first aspect can be dealt with relatively briefly. It is based on the contention that the invention, as claimed in those claims, is not a manner of manufacture because it is a method of treatment of the human body. The patentability under the Act of methods of medical treatment was recently affirmed by the Full Court in Apotex Pty Ltd v Sanofi-Aventis Australia Pty Ltd and Others (No 2) (2012) 204 FCR 494 at [26] and [193]. That decision is binding for the purpose of the present challenge. Nevertheless, it is the subject of an appeal to the High Court which has been heard, with judgment reserved. In these circumstances, Apotex reserved its right to argue that the relevant claims do not claim a manner of manufacture. 362 The second aspect of this challenge to validity relates only to claim 45, which is as follows: The drug as claimed in claim 44, which is contained in a commercial package carrying instructions that the drug should be used for treating schizophrenia, or symptoms thereof. 363 Claim 44 is as follows: A drug when used for treating schizophrenia or symptoms associated with schizophrenia, which comprises Anhydrous Aripiprazole Crystals B defined in any one of claims 12 to 19 in an amount effective to treat schizophrenia or the symptoms thereof, in admixture with a pharmaceutically acceptable diluent. 364 Apotex contended that claim 45 is for a mere collocation. It submitted that there is no working interrelationship between the "commercial package carrying instructions that the drug should be used for treating schizophrenia, or symptoms thereof" and the drug as claimed in claim 44. 365 The question of whether mere collocations constitute patentable subject matter for the purposes of the Act has been recently considered by the Full Court in Smith & Nephew Pty Ltd (ACN 000 087 507) v Wake Forest University Health Sciences and Another (2009) 82 IPR 467. In that case, the relevant claim (claim 49) was for an apparatus for applying negative pressure to a wound beneath a fluid-impermeable seal "wherein said apparatus is in an aseptic package". The Full Court gave detailed consideration to the relevant authorities and, in so doing, drew a distinction between claims containing an impugned integer as an "optional" or "additional" integer, which is merely claim-limiting, and claims where the impugned integer is to be taken, as a matter of construction, to be an essential integer of the combination claimed. The Full Court said (at [32]): [32] We note that the specification does state that the apparatus is preferably packed in a sterile condition to avoid the need for sterilisation of the apparatus prior to use. However, claim 49 has specifically been drawn to include the aseptic package which was asserted before the primary judge to be an essential integer of the combination of the claim. The aseptic package should not in these circumstances be treated as a mere optional addition within the claim or as a limitation on the scope of the claimed combination. Therefore, all the integers of claim 49 are to be considered in determining whether claim 49 is a "mere collocation". 366 Later, the Full Court held (at [43]): [43] In this appeal the respondents have put forward a construction claiming the aseptic package as an essential integer of the combination claimed as the invention. Upon this basis, the only remaining question in this appeal was whether there was a working interrelation brought about by the collocation of integers claimed. On this question, we have found there was not a patentable combination. 367 On its proper construction, claim 45 in the present case claims a combination in which the identified commercial package is to be treated as an essential integer. I accept Apotex's submission that there is no working interrelationship between that packaging and the drug as claimed. 368 The applicants did not advance a detailed submission to the contrary. They merely submitted that this ground of invalidity "should be a ground of last resort" whose "availability and scope … is itself debatable" in light of certain comments made in Lockwood [No 2] at [106]. In my view, that submission does not engage the text of s 18(1)(a) of the Act or the substance of the submission advanced by Apotex. There is nothing in the comments made by the High Court in Lockwood [No 2] at [106] that would deny the availability, in the present case, of Apotex's challenge to claim 45. In my view, that challenge succeeds.

[35]

VALIDITY: CLARITY 369 A claim may be revoked on the ground that it does not comply with s 40(3) of the Act, in that the claim is not clear and succinct: s 138(3)(f) of the Act. 370 This challenge originally related to a number of features of the claims. In final submissions, however, the challenge was limited to two matters. 371 The first matter concerns the definition of certain essential characteristics of Crystals B. Claim 12 requires that Crystals B exhibit an endothermic peak near about 141.5°C in TGA/DTA (heating rate 5°C/min) and an endothermic peak near about 140.7°C in DSC (heating rate 5°C/min). 372 Apotex contended that the claims lacked clarity because "it is not clear whether the endotherms referred to are supposed to be the only endothermic peaks on the respective DSC or TGA curves or whether the claims would encompass multiple peaks, only one of which was close to 140°C". 373 None of the witnesses had any difficulty whatsoever in understanding what claim 12 means when it refers to the respective endothermic peaks exhibited when Crystals B undergo the stipulated TGA/DTA and DSC. The claim simply refers to observable facts. If each identified peak is observed under the stipulated conditions, then those requirements are established. 374 Curiously, Apotex also submitted that its "preferred position" was that the claim is clear and refers to "a unique polymorph, per se". Apotex appeared to relate this submission to a further submission, which was that the Aoki article and the Aoki poster, as well as some of Otsuka's internal documents, indicate that "the appearance of two endotherms suggests a mixture of materials". 375 In cross-examination, Professor Prestidge was asked about the significance to be ascribed to a sample showing two endothermic peaks when undergoing TGA/DTA and DSC. He said: … there are probably two possible explanations for that. Either we start off with a mixture with two forms of a crystalline material that have got different melting points, because that's basically what the endotherm represents, or as we actually ramp up the temperature, we get some sort of phase transformation. 376 In my view, the implications arising from the presence of two endothermic peaks when a sample of material is analysed under each of TGA/DTA and DSC do not bear on the question of whether the claims lack the required clarity. A lack of clarity in the claims does not arise simply because material being analysed might be present with other material as a mixture, and thus give a peak that is characteristic of each material in the mixture when the mixture as a whole is analysed, or because, under the influence of even greater heat, the material undergoes a phase transition after exhibiting an endothermic peak that is characteristic of the material before it undergoes that transition. 377 The second matter raised by Apotex was the requirement in some of the claims that the crystals have "a mean particle size": see claim 16 and claims 30, 31, 35, 36, 43, 44, 45, 111, 112, 119, and 123 to the extent that those claims are dependent directly or indirectly on claim 16. Apotex submitted that these claims are unclear because "the specification of a mean particle size borders on the meaningless in the absence of defined parameters as to how that particle size is to be measured". The vice, according to Apotex, is that the claims are not confined to a particular method of determining "mean particle size". Apotex pointed to evidence that there are different ways of measuring particle size. Before going to that evidence, it is instructive to see what the complete specification says in this regard. 378 When dealing generally with the subject of analytical methods, the complete specification refers to the measurement of particle size as follows: 0.1 g of the particles to be measured were suspended in a 20 ml n-hexane solution of 0.5 g soy lecithin, and particle size was measured using a size distribution meter (Microtrack HRA, Microtrack Co.). 379 The evidence shows that the measurement referred to in this passage involves laser scattering, which measures mean particle size taken as the mean diameter of the particles in question. 380 In his first affidavit, Dr Rowe said that, in September 2001, a number of different techniques were used to measure particle size. These included laser scattering, sieving, Coulter counters, microscopy, and gas absorption. He said that the differences between these techniques can result in different particle size values. 381 In his second affidavit, Dr Rowe said that, in September 2001, he used a Coulter counter, which is designed to measure mean particle size where the particle size is taken as the diameter of the particle. In cross-examination, he also said that he used a Malvern brand laser scattering device to measure particle size, once again taken as the diameter of the particle. In this connection, Dr Rowe volunteered that using a laser scattering device to measure mean particle size was a well-known means of measuring particle size distribution in 2001. 382 With respect to claim 16, Dr Rowe also said in his first affidavit: 87. Having a mean particle size of 50 microns or less means that the average particle size is 50 microns (or less). In my experience, the mean particle size of a powder is (and was in September 2001) usually determined from measurements obtained from different particles in a sample being analysed. Claim 16 does not provide any information relating to the range, or spread, of particle sizes. For example, a powder with a mean particle size of 50 microns may have a range of particle sizes from 45 to 55 microns, or a range from 5 to 200 microns. The range of particle sizes is, in my experience, relevant clinically and in the formulation process, for example, it can influence the properties … 383 Professor Easton also gave evidence that mean particle size can be measured using a variety of methods, including analysis of the refraction and absorption of light by a dispersed solution of the particles being tested. This is the same technique as laser scattering. Professor Easton said that the measurement typically shows the spread of particle sizes which normally follows a bell-shaped curve. He said that it is important to maintain a narrow spread of particle sizes to ensure that all the particles have similar properties. 384 Professor Prestidge gave evidence that the statistical analysis of particle size can be undertaken in various ways. The three ways revealed in his evidence were a volume-weighted mean; a surface-weighted mean; and a volume distribution percentile depending on whether a given particle size is represented in 10%, 50% or 90% of the sample by volume. A volume distribution percentage based on 50% of the volume of the sample (the d(0.5) figure) gives the mean particle size. When questioned on whether one would choose to use the d(0.5) figure as opposed to a volume-weighted mean, Professor Prestidge said that, in his experience, it is more conventional to use the d(0.5) figure. 385 Dr Rowe accepted in cross-examination that, if particle size were to be measured using a laser scattering device (which he accepted was the most common method for measuring mean particle size in 2001), he would assume that the mean particle size would be taken as the d(0.5) figure. 386 In my view, claim 16 and its dependent claims are not unclear with respect to the prescription of a mean particle size. The complete specification makes clear that the analytical method used for determining mean particle size is laser scattering technology. Claim 16 and its dependent claims are to be read in that context. I am satisfied that the person skilled in the art would understand this. The evidence makes clear that there is nothing unusual in relation to the deployment of this technology. Indeed, I am satisfied on the evidence that it was the most common method of determining mean particle size in 2001 and, further, that, for the purpose of analysing a given sample, the person skilled in the art both then and now would understand that, conventionally, the d(0.5) figure would show the mean particle size for the purpose of the relevant claims.

[36]

validity: fair basis 387 A claim may be revoked on the ground that it does not comply with s 40(3) of the Act, in that the claim is not fairly based on the matter described in the specification: s 138(3)(f) of the Act. 388 Apotex contended that claim 16 and its dependent claims are not fairly based because there is no real and reasonably clear disclosure of Crystals B less than 50 microns in size "as part of the invention". Apotex developed this contention by arguing that particle size is not a defining property of Crystals B, which may exist in different particle sizes. It also argued that there is no working interrelationship between crystals having the characteristics of Crystals B and particle size. I have already rejected that contention when dealing with Apotex's challenge based on lack of inventive step: see [358] above. 389 I accept that particle size is not a defining characteristic of Crystals B. It does not follow, however, that claim 16 which adds the limiting feature that the crystals must have a mean particle size of 50 microns or less is, by reason of that limitation, invalid for lack of fair basis. 390 Apotex has not suggested that the added feature of mean particle size is anything but a limitation on the crystals claimed in claim 12. As I discussed in Vehicle Monitoring Systems Pty Ltd (ACN 107 396 136) v Sarb Management Group Pty Ltd (t/as Database Consultants Australia) (ACN 106 549 722) (No 2) (2013) 101 IPR 496 at [121]-[124], fair basis under s 40(3) of the Act deals with the question of claim width. In the present case, claim 12 is wider than claim 16. If given anhydrous aripiprazole crystals infringe claim 16, they will equally and inevitably infringe claim 12. Yet there is no suggestion that claim 12, which covers crystals of any size, is not itself fairly based on the description of the invention in the body of the specification. 391 Regardless of this, the complete specification is replete with references to the milling of aripiprazole crystals and, specifically, the milling of crystals that is required for formulations in which small particle size (stipulated as 50 microns or less) is required. Specific processing steps, including the milling of Hydrate A before final drying, are exemplified for obtaining Crystals B of this size. The complete specification makes clear that, when small particle size is required, the preferred mean particle size is 50 microns or less. 392 In my view, there is a clear and explicit description in the complete specification of the patent of Crystals B having a mean particle size of 50 microns or less. This description alone provides a proper basis for claiming the invention in claim 16.

[37]

Infringement 393 As I have noted, the applicants' case on infringement concerns claims 12, 13, 14, 16, 30, 31, 35, 36, 43, 44, 45, 111, 112, 119, and 123 of the patent. 394 Claim 12 and its dependent claims 13, 14, and 16 claim Crystals B, as therein defined. 395 Claims 30 and 31 are claims to a formulation (the formulation claims). As relevant to this proceeding, claim 30 is dependent directly and claim 31 is dependent indirectly on any one of claims 12, 13, 14 or 16. 396 Claims 35, 36, 111, 112, 119, and 123 are method of treatment claims (the method of treatment claims). As relevant to this proceeding, those claims are dependent on any one of claims 12, 13, 14 or 16. 397 Claim 43, as relevant to this proceeding, is for the use of Crystals B, defined in any of claims 12, 13, 14 or 16, to prepare a medicament to treat or prevent schizophrenia and its symptoms. 398 As relevant to this proceeding, claim 44 is for Crystals B, defined in any one of claims 12, 13, 14 or 16, as a drug in admixture with a pharmaceutically acceptable diluent, to treat schizophrenia and its symptoms. 399 As relevant to this proceeding, claim 45 is directed to the drug described in claim 44, in commercial packaging with instructions that it be used for treating schizophrenia and its symptoms. 400 The threatened conduct is Apotex's sale of the Apotex products in circumstances where those products have been included in the ARTG as suitable for use in treating schizophrenia. The applicants submitted that I should infer that any sale of the Apotex products would be in association with a product information document that would identify those products as suitable for use in treating schizophrenia in accordance with the ARTG registrations. The applicants also submitted that Apotex threatens to infringe the relevant claims by authorising others to exploit the invention as so claimed. They also submitted that Apotex's threatened supply engages s 117 of the Act, which provides: (1) If the use of a product by a person would infringe a patent, the supply of that product by one person to another is an infringement of the patent by the supplier unless the supplier is the patentee or licensee of the patent. (2) A reference in subsection (1) to the use of a product by a person is a reference to: (a) if the product is capable of only one reasonable use, having regard to its nature or design - that use; or (b) if the product is not a staple commercial product - any use of the product, if the supplier had reason to believe that the person would put it to that use; or (c) in any case - the use of the product in accordance with any instructions for the use of the product, or any inducement to use the product, given to the person by the supplier or contained in an advertisement published by or with the authority of the supplier. 401 In this connection, the applicants relied on the circumstances posited by each paragraph of s 117(2) of the Act. 402 Apotex has responded appropriately to these allegations by accepting the following matters: If any of claims 12, 13, and 16 are valid and if, correspondingly, the Apotex products are found to possess the features of those claims, a case of threatened infringement will have been established, which will found the declaratory relief sought in paragraphs 1(a) to (c) of the further amended application and the injunctive relief sought in paragraphs 2(a) to (c) of that application. Although Apotex made no explicit reference to claim 14 in this regard, I am unable to see how it stands outside the concession that has been made. Infringement of the formulation claims would follow a finding of infringement of claims 12, 13, and 14. Although Apotex made no explicit reference to claim 16 in this regard, I am unable to see how that claim stands outside the concession that has been made. Apotex was also content for the infringement of claim 44 to be treated in the same way. If the Apotex products are found to possess the features of claims 12, 13, and 16, s 117(2)(c) of the Act will be engaged, based on the underlying infringement of claims 35 and 36. Apotex accepted that the case based on the underlying infringement of claims 111, 112, and 119 should be treated in the same way. Once again, I am unable to see how claim 14 falls outside the concessions that have been made. If claim 45 is valid, its infringement should be treated in the same way as the infringement of the formulation claims. 403 Apotex submitted that claim 123 is meaningless and provides no additional basis for infringement beyond claims 35 and 36. 404 Apotex's defence focused on what it submitted was the proper construction of claim 12 of the patent relating to the characteristic endothermic peaks exhibited in TGA/DTA and DSC. It submitted that, properly construed, claim 12 specified, for Crystals B, the presence of one peak only in TGA/DTA and one peak only in DSC. The significance of this submission lies in the fact that the applicants' evidence, led through Professor Prestidge, was that, on material supplied by Apotex, samples of the Apotex products do not exhibit one endothermic peak in TGA/DTA and one endothermic peak in DSC but, in each case, two endothermic peaks. Professor Prestidge's evidence was that, in the case of TGA/DTA, one of the peaks is near about 141.5°C and, in the case of DSC, one of the peaks is near about 140.7°C. Apotex argued that, because the Apotex products exhibit, in each form of analysis, two peaks, those products fall outside claim 12. No issue appeared to be taken by Apotex regarding the presence in the samples of the other characteristics of Crystals B as defined in claim 12 or the presence of a mean particle size falling within the feature added by claim 16. 405 This part of Apotex's defence is answered by my findings on the issue of lack of clarity: see [376] above. It is no answer to the case on infringement that Crystals B might be present with another substance showing the other endothermic peak in each analysis or that, assuming only Crystals B to be present, they undergo a phase transition, under the influence of even greater heat, after exhibiting an endothermic peak that is characteristic of Crystals B before they undergo that transition. The effect of Professor Prestidge's evidence is that the physicochemical properties reported in the documents provided by Apotex for the active pharmaceutical ingredient in the Apotex products are consistent with the characteristics of Crystals B defined in claims 12, 13, 14, and 16. I am satisfied on the balance of probabilities that the Apotex products contain Crystals B, as so defined. 406 I turn to Apotex's remaining arguments. 407 I am not persuaded that claim 123 is meaningless. Ultimately, Apotex did not seek its revocation on the ground that it was meaningless. In any event, as matters have transpired, it is not necessary for the applicants to rely on claim 123 in order to establish infringement, given Apotex's concessions to which I have referred. 408 I am not persuaded that the applicants have established infringement by Apotex of claim 43, which is plainly directed to the preparation of a medicament. There is no evidence that Apotex itself has prepared the medicament forming part of the Apotex products (being the finished products supplied by Apotex in the patent area). It is equally probable that Apotex is no more than a purchaser of a medicament manufactured independently by another person which is then used by Apotex for the Apotex products. Moreover, there is no evidence about where any relevant medicament is prepared. 409 I am persuaded that the applicants have established a case of threatened authorisation by Apotex in relation to the method of treatment claims. In its written submissions, Apotex appeared to place reliance on the non-exhaustive factors specified in s 101(1A) of the Copyright Act 1968 (Cth) (the Copyright Act) concerning the infringement of copyright by authorisation in subject matter other than works. That provision was introduced into the Copyright Act by the Copyright Amendment (Digital Agenda) Act 2000 (Cth). I am unable to see the direct relevance of that provision to s 13(1) of the Act, which has no counterpart to s 101(1A) of the Copyright Act. Nevertheless, I will proceed on the basis that "authorise" in s 13(1) was intended to have the judicially-received meaning of that word in the context of the Copyright Act as at 1990: Bristol-Myers Squibb Company v F H Faulding & Co Limited (2000) 97 FCR 524 at [97]. In The University of New South Wales v Moorhouse and Another (1975) 133 CLR 1 at 12-13, Gibbs J (as he then was) accepted, in the copyright context, that "authorise" had the meaning of "sanction, approve, countenance". His Honour went on to say that a person who has under his or her control the means by which an infringement of copyright may be committed, and makes that means available to other persons knowing or having reason to suspect that it is likely that the means will be used to commit an infringement, would authorise any infringement that resulted from use of the means, if the person failed to take reasonable steps to limit its use to a legitimate purpose. In the present case, there is no doubt that Apotex intends to supply the Apotex products so that they will be used to treat schizophrenia. In those circumstances, I am unable to see how Apotex's threatened conduct, if carried out, would not be an authorisation of others to use the Apotex products according to the method of treatment claims. 410 For completeness, I should record that I am satisfied that s 117(1) is also engaged in the present case by s 117(2)(a) and s 117(2)(b) of the Act. 411 I am uncertain as to whether the evidence establishes infringement of claim 14. The parties did not address this claim in isolation from claim 12. It is apparent that the two claims are different. I wish to hear the parties further on that question if, in view of the other findings I make, a case of infringement based on claim 14 is to be pursued. 412 Subject to that matter, I am satisfied that, other than in relation to claim 43, and claim 45 which I have found to be invalid, the applicants have established their case on infringement.

[38]

Exclusive licence 413 There remains a question of BMS's standing to sue for infringement. This depends on it being an exclusive licensee: s 120(1) of the Act. An exclusive licensee for the purposes of the Act is defined to mean: … 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. 414 The Act defines the word "exploit" inclusively: exploit, in relation to an invention, includes: (a) where the invention is a product - make, hire, sell or otherwise dispose of the product, offer to make, sell, hire or otherwise dispose of it, use or import it, or keep it for the purpose of doing any of those things; or (b) where the invention is a method or process - use the method or process or do any act mentioned in paragraph (a) in respect of a product resulting from such use. 415 BMS has been entered in the Register of Patents as a licensee of the patent. Apotex accepts that BMS is a licensee. It does not accept, however, that BMS is the exclusive licensee. 416 BMS bases its claimed entitlement on the legal effect of an agreement between Otsuka and BMS styled as a Restated Development and Commercialization Collaboration Agreement made on 23 October 2001 (the Agreement). The Agreement has been amended on a number of occasions, but these amendments are not material to the present question. The Agreement relates to the arrangement between Otsuka and BMS regarding the development and commercialisation of aripiprazole worldwide. It is a prolix document. 417 Clause 2 of the Agreement contains a lengthy recitation of the background to and intended scope of the collaboration between Otsuka and BMS regarding "the Product" which, for present purposes, can be taken to be any prescription, finished, human pharmaceutical product in any formulation containing aripiprazole in any compound form. Despite its length, this recital is instructive as to the relationship between the applicants so far as the patent is concerned. Clause 2 of the Agreement provides: Through years of extensive and costly research and development, Otsuka has developed the Compound, which it believes can be developed into Product having considerable worldwide commercial and therapeutic potential. Recognizing the high cost and complexity of necessary further development, and the significant accompanying commercial risks, Otsuka has concluded that, in order to achieve the full potential of the Compound and Product, it is necessary to collaborate with another company. BMS has considerable experience developing, obtaining regulatory approval of and commercializing prescription drugs worldwide, and BMS shares Otsuka's views as to the commercial and therapeutic potential of the Compound and Product. Accordingly, as provided in this Agreement, Otsuka and BMS agree to collaborate on the continued development and Commercialization of Product. The parties believe this collaboration, on the terms and conditions set forth in this Agreement, is essential to achieve the goal of completing the development of Product and making it available to customers throughout the Territory at the earliest possible time. The parties hope to achieve commercial success through this collaboration, and they also believe that this collaboration will be to the benefit of human health worldwide. As described in detail in this Agreement, Otsuka hereby retains BMS, as a contract service provider on behalf of Otsuka in the United States and the European Union, to continue the development of Product in collaboration with Otsuka and to Commercialize the Product in collaboration with Otsuka. Otsuka or its Affiliates will be the holder of the Product NDA in the United States and the MA(s) in the European Union, and, to the maximum extent permissible, BMS will Commercialize Product in the United States and the European Union on behalf of Otsuka, under Otsuka Trademark(s), recording Product sales in the name of Otsuka or its Affiliates. Otsuka retains the option to Co-Promote Product in the United States and the other Co-Promotion Countries in the European Union. Otsuka also retains the Reserved Territory for its Commercialization. As provided below, BMS has an exclusive license to Commercialize Product in the Rest of Territory (countries in the Territory outside the United States and the European Union). 418 There is no doubt that the territory over which this "exclusive license" can be exercised includes the patent area defined for the purposes of the Act. The word "Commercialize" is defined as "to engage" in the following activities: … advertising, marketing, promotion, sale and distribution of a product, and activities related thereto … 419 Clause 5.4.1 of the Agreement contains the grant of an "exclusive license" that is subject to reservations. The "exclusive license" is stated to be: … under the Patent Rights and Otsuka's related know-how, data and information, to make and have made (subject to Section 5.4.3(c)), use and sell Product, under [certain trade marks], and to formulate [aripiprazole] into Bulk Tablets and other Product, in the Field in the Rest of Territory. For the sake of clarification … the Field for Product containing a Related Compound is limited to Neuroscience Indications. … 420 It is not necessary to set out all the defined terms. It is only necessary to discuss the term "Patent Rights". This term is defined by reference to certain listed patents, which do not include the patent. However, the meaning of the term is extended to: … all other patents and patent applications (and patents issuing from such applications) that become owned, solely or jointly, by Otsuka … which generically or specifically relate to [aripiprazole] … 421 The reservations exclude from the licence the right to manufacture aripiprazole in its various forms: clause 5.4.1(a). Somewhat redundantly, clause 5.4.3(c) provides: Notwithstanding the provisions of Sections 5.4.1 and 5.4.2, Otsuka reserves the exclusive worldwide right, for itself … to manufacture, or have manufactured [aripiprazole in its various forms] … 422 The point of present significance is that the licence granted under the Agreement to BMS does not include the right to manufacture aripiprazole under one of Otsuka's patents. There are other reservations and exclusions, but these do not concern the patent or the exercise of rights in the patent area. 423 In oral submissions, Apotex pointed to two matters which, it argued, signified that the Agreement did not confer on BMS an exclusive licence with respect to the patent. The first matter is that, whereas the right to "make" is one of the stipulated activities falling within the meaning of "exploit" as defined for the purposes of the Act, BMS does not have a licence to manufacture aripiprazole in its various forms. Apotex argued that the exclusion of this right from the licence, and its correlative reservation to Otsuka, necessarily meant that the Agreement does not confer an exclusive licence within the meaning of the Act. The second matter is that, according to Apotex, it is not clear on the face of the Agreement that the rights it does confer on BMS are exclusive of Otsuka's own rights to engage in the same activities with respect to aripiprazole. In short, Apotex argued that it is not clear that exclusive, as opposed to sole, rights are conferred by the Agreement. 424 The first matter raises a difficult question of statutory construction. 425 The Patents Act 1952 (Cth) (the 1952 Act) recognised the rights of an exclusive licensee. Section 6 of the 1952 Act defined the expression "exclusive licensee" as follows: … a licensee under a licence granted by the patentee which confers on the licensee, or on the licensee and persons authorized by him, the right to make, use, exercise and vend the patented invention, throughout Australia, to the exclusion of all other persons, including the patentee… 426 This definition was discussed by the High Court in Ex parte British Nylon Spinners Limited and Imperial Chemical Industries Limited; In re Imperial Chemical Industries Limited's Patent (1963) 109 CLR 336. In that case, the applicants applied to extend the term of a patent relating to an improved process for melt-spinning Nylon yarn. The right to apply for an extension of term was conferred by the 1952 Act on the patentee who, by dint of s 96, was taken to include "an exclusive licensee". The patentee in that case had granted British Nylon Spinners Limited (BNS) what was called an exclusive licence to make, use, exercise, and vend the invention within a certain limited field (namely, yarn of which the filament, or any of the filaments, do not exceed 0.09 mm in its largest cross-sectional dimension in the drawn condition). The patentee later granted Imperial Chemical Industries Limited (ICI) a licence under the patent which was expressed to "be subject to the B.N.S. licence but otherwise exclusive". Both BNS and ICI claimed to be entitled to apply to extend the term of the process patent. They argued that each licence, within its limited field, conferred a right to make, use, exercise, and vend the patented invention throughout Australia "to the exclusion of all other persons, including the patentee". 427 The High Court emphatically rejected that argument, branding it "fallacious in the extreme". The High Court said (at 340): In approaching this question it is to be observed that any number of licences may be granted conferring a right to make use exercise and vend a patented invention. Perhaps it may be said that any number of licences may be granted conferring the right to make use exercise and vend a patented invention. But the introduction of the definite article into the definition suggests that the right must be found in a single licensee for it is the right to make use exercise and vend the patented invention which the licence must be found to confer. However this may be, the concluding words of the definition make it clear beyond doubt that the licence must confer that right to the exclusion of all other persons including the patentee. One of several licences which confer upon each of the respective licensees a right to make use exercise and vend a patented invention within a series of limited fields does not, therefore, constitute an exclusive licence. It may be true to say that each of such licensees has the exclusive right within his limited field to make use exercise and vend the patented invention but this is far from saying that each has the right to make use exercise and vend the patented invention throughout Australia to the exclusion of each other. 428 In dealing with that question, the High Court observed that s 101 of the Patents Act 1949 (UK) contained a definition of "exclusive licensee" that was substantially different to that used in the 1952 Act. The definition in the UK Act recognised as an exclusive licensee a licensee of any right in respect of the patented invention that was to the exclusion of all other persons, including the patentee. It had been held that that definition permitted "a plurality of exclusive licensees to be created in respect of any one patent monopoly": In the Matter of Courtaulds Ld.'s Application for Extension of the Term of Letters Patent No. 511,160 [1956] RPC 208 at 210. Plainly, by drawing attention to that decision, the High Court was distinguishing the position under the 1952 Act, where there could be only one exclusive licensee, not a multiplicity of such licensees. 429 In Uprising Dragon Ltd and Another v Benedict Trading & Shipping Pty Ltd and Others (1987) 16 FCR 93 at 102, French J (when in this Court) saw British Nylon Spinners as deciding that an exclusive licence limited to a particular field will not amount to an exclusive licence for the purposes of s 6 of the 1952 Act. 430 The meaning of "exclusive licensee" in the Act arose for consideration by the Supreme Court of Queensland in Grant and Another v Australian Temporary Fencing Pty Ltd (2003) 59 IPR 170. In that case, the defendant, who had been sued for patent infringement, challenged an assignment to the first plaintiff of a patent (for portable fencing used in roadworks) and the consequent grant by the first plaintiff of an exclusive licence of the patent to the second plaintiff. The defendant challenged the existence of an exclusive licence on a number of bases, including that the licence conferred no right to import the patented product, only the right to use it. It was argued that, as the licence did not grant all the rights specified in the Act as constituting the "right to exploit", there was no grant of an exclusive licence. This argument was rejected by Holmes J. After noting the terms of s 6 of the 1952 Act, her Honour (at [41]) said: [41] It can be seen that the definition in the 1952 Act is exhaustive, and the rights cumulative: making, using, exercising and vending. That is in contrast with the non-exhaustive definition of "exclusive licensee" under the present legislation, which merely refers to a "right to exploit" and then separately defines "exploit" in an inclusive and distributive way: "make, hire, sell or otherwise dispose … use or import …" (emphasis added). The latter definition is at least open to a construction similar to that given [to] s 101 of the English Patents Act of 1949, as discussed in Ex Parte British Nylon Spinners Ltd and Imperial Chemical Industries Ltd: that because the definition of exclusive licence in that provision referred to a licence conferring "any right in respect of the patented invention" the way was open for a "plurality of exclusive licences". But more importantly for present purposes, it makes no sense to say that the rights explicitly conferred by a licence must, in order to be exclusive, be an exhaustive list of what is comprised in the right to exploit in the legislation, when the legislation itself is not exhaustive in its definition of the term. Consequently, I reject the argument that the failure to include a right to import (which neither party may have contemplated exercising, before or after the grant of licence) is fatal to exclusivity. [Footnotes omitted] 431 In Pharmacia Italia SpA and Another v Interpharma Pty Ltd (2005) 67 IPR 397, Sundberg J, when dealing with an application for an interlocutory injunction for patent infringement, referred to the decision in Grant. At [21], his Honour observed that the argument that a licence could not be an exclusive licence if it did not confer the right to import was rightly rejected. 432 The applicants contended that I should follow Grant and Interpharma with respect to the meaning of "exclusive licensee" as used in the Act. They submitted that the licence conferred by the Agreement should not be denied the status of an exclusive licence simply because it does not confer on BMS the right to manufacture aripiprazole. 433 I am, with respect, unable to agree with the reasoning and conclusion in Grant. 434 Section 13(1) of the Act provides that a patent gives the patentee the exclusive rights to exploit the invention (which can only mean the invention as claimed) and to authorise another person to exploit the invention. The definition of "exclusive licensee" recognises these twin rights. The licence referred to in the definition might confer on the licensee the patentee's exclusive right to exploit the patented invention throughout the patent area. Alternatively, the licence might confer that right on the licensee and persons authorised by the licensee. In either case, 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. It follows that the legal effect of the licence must also be to exclude the patentee from exercising the exclusive right under s 13(1) of the Act to authorise another person to exploit the invention. 435 In this way, it can be seen that the definition of "exclusive licensee" operates in harmony with the rights conferred by the Act on the patentee. The licence under which the licensee derives the status of "exclusive" licensee is one that excludes the patentee, and others deriving authority from the patentee, from exercising the rights conferred by the patent with respect to the invention during the term of the licence. An exclusive licence cannot be one that reserves to the patentee or any other person any residual right with respect to the exploitation of the invention. It follows that there can only be one exclusive licensee. 436 In my view, the Act speaks of "the right to exploit" the invention as a single, indivisible right. The word "exploit" is used in the Act as a hypernym to cover a range of activities. The inclusive nature of the definition employed by the Act, and the exemplification of particular activities to elucidate the meaning of the word, signify that a broad range of activities is intended to be covered. However, s 13(1) of the Act recognises that only two rights are conferred by the patent: the right to exploit the invention and the right to authorise others to exploit the invention. The use of disjunctive language in the definition of "exploit" to identify particular activities falling within the scope of the term does not create separate rights with respect to those activities. It merely recognises that the right to exploit covers a range of activities, any one of which, if undertaken, would amount to the exercise of the right to exploit. 437 Support for this view is provided by the provisions in the Act respecting assignment. Section 13(2) of the Act recognises that the exclusive right to exploit an invention, as an incident of the grant of a patent for that invention, is personal property that is capable of assignment and of devolution by law. Section 14(2) of the Act recognises that a patent may be assigned for a place in, or part of, the patent area. There is no recognition, however, that a patent might be assigned as to some incident of the right to exploit the invention. 438 This reasoning reflects, in my respectful view, the essential reasoning of the High Court in British Nylon Spinners with respect to the meaning of "exclusive licensee" in s 6 of the 1952 Act: the patentee has conferred on the licensee, exclusively, the exercise of the rights that the patentee itself has been granted under the patent. I am unable to see how, in determining the meaning of "exclusive licensee" in the present Act, this reasoning can be avoided by distinguishing the definition of the word "exploit" from the words "make, use, exercise and vend" in the corresponding definition of "exclusive licensee" in the 1952 Act. There are plainly differences between the two definitions in language and structure, including that the definition under the 1952 Act is exhaustive whereas the definition under the Act is merely inclusive. But, in my respectful view, these differences do not have the significance attributed to them in Grant. Regardless of possible differences in scope between "exploit" under the Act and "make, use, exercise and vend" under the 1952 Act, the definition of "exclusive licensee" in each Act refers to the conferral by the patentee of a single licence that precludes the patentee, and any person deriving authority from the patentee, from exercising the rights granted by the patent. 439 Given that the Agreement excludes from the licence granted to BMS the right to manufacture aripiprazole (including in the form of Crystals B), and reserves that right to Otsuka, I do not accept that BMS is an exclusive licensee of the patent with title to sue Apotex for infringement. 440 In these circumstances, it is unnecessary for me to decide whether the Agreement confers exclusive rights or merely sole rights on BMS. Nevertheless, appreciating that a different view might be taken on the question I have decidedly adversely to BMS, I should state that, in respect of the rights in the Agreement that are expressed to be granted to BMS exclusively, there is nothing that indicates to me that only sole rights were intended to be granted.

[41]

Schedule A 13 Anhydrous Aripiprazole Crystals B according to claim 12, wherein said low hygroscopicity is a moisture content of 0.10% or less after placing said drug substance for 24 hours in a closed container maintained at a temperature of 60°C and a humidity level of 100%. 14 Anhydrous Aripiprazole Crystals B according to claim 12, having a powder x-ray diffraction spectrum having characteristic peaks at 20 = 11.0°, 16.6°, 19.3°, 20.3° and 22.1°. 16 Anhydrous Aripiprazole Crystals B according to claim 12, wherein said crystals have a mean particle size of 50 µm or less. 30 The Anhydrous Aripiprazole Crystals B according to any one of claims 12 to 17 formulated with one or more pharmaceutically acceptable carriers. 31 The Anhydrous Aripiprazole Crystals B according to claim 30 formulated with one or more pharmaceutically acceptable carriers to form a solid oral tablet. 35 The use of Anhydrous Aripiprazole Crystals B defined in any one of claims 12 to 19 for the treatment of a central nervous system disorder. 36 The use of Anhydrous Aripiprazole Crystals B defined in any one of claims 12 to 19 for the treatment of schizophrenia. 43 The use of Anhydrous Aripiprazole Crystals B defined in any one of claims 12 to 19 to prepare a medicament to treat or prevent schizophrenia and the symptoms associated with schizophrenia. 44 A drug when used for treating schizophrenia or symptoms associated with schizophrenia, which comprises Anhydrous Aripiprazole Crystals B defined in any one of claims 12 to 19 in an amount effective to treat schizophrenia or the symptoms thereof, in admixture with a pharmaceutically acceptable diluent. 45 The drug as claimed in claim 44, which is contained in a commercial package carrying instructions that the drug should be used for treating schizophrenia, or symptoms thereof. 111 A method of treating a central nervous system disorder comprising administering Anhydrous Aripiprazole Crystals B defined in any one of claims 12 to 19. 112 A method of treating schizophrenia comprising administering Anhydrous Aripiprazole Crystals B defined in any one of claims 12 to 19. 119 A method of treating or preventing schizophrenia and symptoms associated with schizophrenia comprising administering Anhydrous Aripiprazole Crystals B defined in any one of claims 12 to 19. 123 Use according to any one of claims 35-43, substantially as hereinbefore described.

Parties

Applicant/Plaintiff:

Bristol-Myers Squibb Company

Respondent/Defendant:

Apotex Pty Ltd

Legislation Cited (5)

National Health Act 1953(Cth)

Patents Act 1952(Cth)ss 6, 96

Patents Act 1990(Cth)ss 7, 13, 14, 18, 40, 117, 120(1), 138(3)

Cases Cited (21)

l Chemical Industries Limited; In re Imperial Chemical Industries Limited's Patent (1963) 109 CLR 336 Grant and Another v Australian Temporary Fencing Pty Ltd (2003) 59 IPR 170 In the Matter of Courtaulds Ld.'s Application for Extension of the Term of Letters Patent No. 511,160 [1956] RPC 208 Lockwood Security Products Pty Ltd v Doric Products Pty Ltd [No 2] (2007) 235 CLR 173 Pharmacia Italia SpA and Another v Interpharma Pty Ltd (2005) 67 IPR 397 Prestige Group (Australia) Pty Ltd v Dart Industries Inc (1990) 26 FCR 197 Sanofi-Aventis Australia Pty Ltd and Others v Apotex Pty Ltd (No 3) (2011) 196 FCR 1 Sigma Pharmaceuticals (Australia) Pty Ltd v Wyeth and Another (2010) 88 IPR 459 Smith & Nephew Pty Ltd (ACN 000 087 507) v Wake Forest University Health Sciences and Another (2009) 82 IPR 467 The General Tire & Rubber Company v. The Firestone Tyre and Rubber Company Limited and Others [1972] RPC 457 The University of New South Wales v Moorhouse and Another (1975) 133 CLR 1 The Wellcome Foundation Limited v V.R. Laboratories (Aust.) Proprietary Limited (1981) 148 CLR 262 Uprising Dragon Ltd and Another v Benedict Trading & Shipping Pty Ltd and Others (1987) 16 FCR 93 Vehicle Monitoring Systems Pty Ltd (ACN 107 396 136) v Sarb Management Group Pty Ltd (t/as Database Consultants Australia) (ACN 106 549 722) (No 2) (2013) 101 IPR 496 Wake Forest University Health Sciences and Others v Smith & Nephew Pty Ltd (ACN 000 087 507) and Others (No 2) (2011) 92 IPR 496

(2002) 212 CLR 411

(2008) 76 IPR 618

(2009) 82 IPR 416

(2013) 100 IPR 285

(2012) 204 FCR 494

(2000) 97 FCR 524

(1963) 109 CLR 336

(2003) 59 IPR 170

(2007) 235 CLR 173

(2005) 67 IPR 397

(1990) 26 FCR 197

(2011) 196 FCR 1

(2010) 88 IPR 459

(2009) 82 IPR 467

(1975) 133 CLR 1

(1981) 148 CLR 262

(1987) 16 FCR 93

(2013) 101 IPR 496

(2011) 92 IPR 496

(2010) 87 IPR 516

Cited By (3)

Actavis Pty Ltd v Orion Corporation

Bristol-Myers Squibb Company v Apotex Pty Ltd