NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GmbH; NAPP PHARMACEUTICAL HOLDINGS LTD v SANDOZ LTD

Transcription

1 [2009] R.P.C NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GmbH; NAPP PHARMACEUTICAL HOLDINGS LTD v SANDOZ LTD COURT OF APPEAL H1 (Jacob L.J., Lord Neuberger and Lawrence Collins L.J.): March and 1 April, 2009 [2009] EWCA Civ 252, [2009] R.P.C. 18 Patent Sustained release formulation of oxycodone Infringement Validity Claims held valid but not infringed Appeal to Court of Appeal Construction of claims Meaning of spheronising agent Whether release coating had to control release of all oxycodone in formulation Added matter Divisional patents Whether disclaimer added matter Permissible disclaimers Lack of Inventive step Whether obvious over prior art Whether oxycodone would have been perceived as an alternative to morphine at priority date Whether motive to find solution to problem relevant Whether unexpected advantage failed to defeat obviousness attack H2 European Patent Convention, Art.123(2) H3 The respondent/claimant in each action sought to revoke European Patent No and European Patent No in the name of the appellant/defendant and also sought a declaration of non-infringement in respect of formulations which each wished to market in the United Kingdom. The appellant counterclaimed that the patents were valid and infringed. The declaratory proceedings thus fell away and the actions proceeded as conventional patent actions brought by the appellant against each of the respondents. H4 Each of the patents was concerned with a sustained release formulation of oxycodone, an opioid pain killer. Oxycodone had been discovered in 1916 and was in clinical use since The respondents asserted that each of the patents was invalid on the basis of added matter and on the basis of lack of inventive step over the common general knowledge and over two patents referred to as Oshlack and Goldie. Certain of the added matter attacks were met by an application to amend, which was not opposed. The application to amend was conditional in the case of 730 but unconditional in respect of 246. The amendments to 246 did not affect the scope of the claims. H5 The appellant asserted that claims 6, 7, 9, 11 to 14 of 730 and claims 1, 4, 9, 16 to 19 of 246 were infringed by a product manufactured by Acino AG (then Cimex AG) which each of the respondents intended to market in the United Kingdom. This product contained, embedded in excipients, small particles made up of a number of layers. 1 The core was bought in by Cimex and was made from sugar crystals which were built on with liquid glucose and corn starch in a standard coating vessel from which they were taken and dried. The inner layer, which surrounded the core, was made up of oxycodone hydrochloride (83 wt%), hydroxypropylmethyl cellulose (HPMC, about 8 wt%) and talc/macrogol (about 9 wt%). It was sprayed onto the cores in a 1 A diagram of a cross-section of a particle showing the layers is set out in the judgment at [20].

2 H6 H7 540 NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GMBH fluid bed processor. The polymer layer, which surrounded the inner layer, contained ethyl cellulose, hydroxypropylcellulose and propylene glycol. It was accepted by the respondents that the polymer layer controlled the release of the oxycodone hydrochloride in the inner layer. The final and outer layer, which surrounded the polymer layer, contained about 96 wt% oxycodone hydrochloride and about 4 wt% HPMC. About 20% of the total weight of oxycodone hydrochloride in the dosage form was in the outer layer. Both patents had a priority date of 1991 and 246 was a divisional of 730 which in turn was a divisional from the original PCT application. Claims 1 and 6 of 730 were as follows: 1. A controlled release oxycodone formulation for oral administration to human patients, comprising: (a) oxycodone salt in an amount equivalent to 10 mg to 160 mg of the oxycodone hydrochloride salt, and (b) a controlled release dosage matrix, other than an acrylic resin matrix selected so that the formulation provides ph-independent dissolution characteristics, (c) wherein said formulation provides, at steady state after repeated administration at 12-hour intervals, a mean maximum plasma concentration of oxycodone of 6 to 240 ng/ml at 2 to 4.5 hours after administration and a mean minimum plasma concentration of oxycodone of 3 to 120 ng/ml at 10 to 14 hours after administration. 6. A controlled release oxycodone formulation for administration to human patients, comprising: (a) an analgesically effective amount of spheroids comprising oxycodone or a salt thereof and a spheronising agent; (b) each spheroid being coated with a film coating which controls the release of the oxycodone or oxycodone salt at a controlled rate in an aqueous medium. Claim 1 of 246 was as follows: 1. A controlled release oxycodone dosage form for oral administration to human patients, comprising: (a) oxycodone salt in an amount equivalent to 10 mg to 160 mg of the oxycodone hydrochloride salt; (b) a matrix incorporating said oxycodone salt; (c) a coating on said matrix controlling the release of said oxycodone salt; (d) wherein said dosage form has an in vitro dissolution rate, when measured by the USP Paddle Method at 100 rpm in 900 ml aqueous buffer (ph between 1.6 and 7.2) at 37 C, between 12.5% and 42.5% (by weight) oxycodone salt released after 1 hour, between 25% and 55% (by weight) oxycodone salt released after 2 hours, between 45% and 75% (by weight) oxycodone salt released after 4 hours and between 55% and 85% (by weight) oxycodone salt released after 6 hours. Published by Oxford University Press for the Intellectual Property Office

3 H8 H9 H10 H11 H12 H13 H14 H15 [2009] R.P.C The Patents Court held that the patents were both valid but not infringed by the Cimex product. 2 The patentee/defendant appealed against the finding on non-infringement and the claimants appealed against the finding of validity. In relation to claim 6 of 730, the respondents had argued before the Patents Court inter alia that the term spheronising agent, in its purest form, referred to materials such as microcrystalline cellulose, which were incorporated to give the material the correct degree of plasticity in an extrusion/spheronisation process. They submitted that, at its widest, the term was limited to agents which had assisted in making a sphere out of something that was not a sphere. The appellant submitted that a spheronising agent was anything that would assist in the process of making a spheroids and claimed that the HPMC in the inner layer of the Cimex product was the spheronising agent. The judge held that claim 6 of 730 used the term spheronising agent in the sense of an agent which had enabled the excipients to form into a sphere and that the HPMC in the inner layer was not a spheronising agent within the meaning of claim 6 of 730. In relation to claim 1 of 246 the respondents had argued before the Patents Court that the claim required all the oxycodone salt in the formulation to be in the film coating whereas in the Cimex formulation 20% of the oxycodone was in the outer layer and was released immediately, with the release of the remaining 80% of the oxycodone being controlled by the film coating. The appellant had submitted that the claim requirement was satisfied if the coating controlled the release of such oxycodone as was beneath the coating. The judge upheld the respondents contention that claim 1 of 246 required all the oxycodone salt in the formulation to be in the film coating. In addition to a conventional attack on added matter, the respondents had pleaded that the matter disclosed in the specification of 730 extended beyond that disclosed in the parent PCT Application WO 93/10765 filed on 25 November 1992 and, in particular, that 730 contained an undisclosed disclaimer which was prohibited by the case law of the EPO. 3 The judge rejected the respondents argument on added matter and said that the filing of a divisional application with a disclaimer which was not based on the original disclosure was not analogous to the cases where the Enlarged Board had held that disclaimers might be permissible. He also held that the test for added matter was that set out in the European Patents Convention and the Patents Act On appeal, the respondents renewed their argument based on undisclosed disclaimers. The respondents had argued before the Patents Court that the PCT application disclosed that acrylic resins were suitable materials for inclusion in a controlled release matrix formulation of the invention; that 730 patent disclosed that the matrix may be any matrix...except an acrylic resin matrix being selected so that the formulation provides ph independent dissolution characteristics ; that claim 1 of 730 claimed a controlled release dosage matrix, other than an acrylic resin matrix selected so that the formulation provides ph-independent dissolution characteristics which amounted to a disclaimer which had not been disclosed in the PCT application. The judge rejected the attack based on added matter. On appeal, the respondents argued that process of division has created a disclosure of two alternative classes of acrylic resin which could be used to produce an overall ph-independent formulation an acrylic resin which was itself ph-independent and one 2 Ratiopharm GmbH v Napp Pharmaceutical Holdings Ltd; Sandoz Ltd v Napp Pharmaceutical Holdings Ltd [2009] R.P.C. 11, Pat Ct. 3 Case G001/03, PPG Industries Ohio, Inc 8 April 2004.

4 H16 H17 H18 H19 H NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GMBH which was not; and that it would have been all right to have divided out these two classes if they had been separately disclosed in the application, but they were not. In the Patents Court, the judge rejected the attack based on lack of inventive step over the common general knowledge, Oshlack and Goldie. On the appeal, the respondents pursued only the attack based on Oshlack. In relation to common general knowledge, the judge had found that the evidence established that before the priority date oxycodone was generally perceived as a weak opioid analgesic and relatively little was known about it as a single entity oral dosage form in The majority of those clinicians who had had experience of oxycodone would have regarded it as a co-drug administered with aspirin or paracetamol to give additional analgesia by another mechanism. They would have known that it was a potent opioid but most clinicians would have had no practical experience of the use of oxycodone as an opiod alternative to morphine in 1991 and that such use did not form part of the common general knowledge. The judge rejected the respondents expert s evidence that oxycodone would have been included in a short list of known drugs as suitable alternatives for controlled release morphine. Oshlack was entitled Controlled release bases for pharmaceuticals and had a publication date of 20 January It disclosed that by using a combination of a higher aliphatic alcohol and an acrylic resin, the release of an active ingredient from a controlled release matrix could be varied over a period of time from 5 to 24 hours. Oxycodone was identified as being a narcotic analgesic and a suitable drug for slow release. The judge held that compared with the enabled disclosure of Oshlack, the differences with the inventive concepts of the various claims in issue were that each claim relied on required at least one of the following features: (i) the provision at steady state after 12 hourly dosing of maximum and minimum plasma concentrations within ranges; (ii) a film coated spheroid formulation; (iii) an in vitro release rate by USP Paddle Method (still releasing after 4 hours); and (iv) a coating controlling release. He held that Oshlack was a document aimed at a formulator and that its disclosure was concerned far more with the novel controlled release matrix than with the identity of the drugs released by it; that it was not the sort of document which any of the clinicians who had given evidence would have been accustomed to reading or which would have much influenced their thinking and neither was it a document which would have influenced the thinking of a pharmacologist. He held that the skilled team would not have been sufficiently motivated by Oshlack either alone or in the light of common general knowledge to have embarked on a development programme so as to arrive at a 12 hourly controlled release dosage form of oxycodone. On appeal, the respondents argued that having been presented by Oshlack with the concept of a controlled release formulation of oxycodone, the clinical issue (whether there was any purpose in making up such a formulation) disappeared; that in terms of problem and solution, the problem was no longer can you find an alternative to morphine (the clinical question) but can you make up a formulation of oxycodone suitable for 12 hourly administration given the teaching of Oshlack (a pure formulation question) and that that was why the clinician s motivation, crucial to assessing obviousness over common general knowledge, was irrelevant when considering obviousness over Oshlack. They submitted that the formulator, if he had been asked the second question, would have said yes without the need for experimentation because he could have predicted with reasonable certainty that he could so and that, accordingly, in terms of problem and solution, the solution was obvious. They contended that the difference from obviousness over common general knowledge was that Oshlack Published by Oxford University Press for the Intellectual Property Office

5 [2009] R.P.C H21 H22 H23 H24 H25 H26 H27 asked the formulator the question and the formulator s expertise gave the answer; that obviousness over common general knowledge required the clinician to ask the formulator the question and, on the facts, that would not have happened so the answer would never have been sought. They submitted that was why motivation was central to the attack based on obviousness but was irrelevant to the case based on Oshlack. The respondents also argued that the alleged invention was a case of an unexpected advantage of the sort not found to save a patent from an obviousness attack. 4 Held, allowing the appeal on infringement and dismissing the cross-appeal on validity and finding the patents valid and infringed: Infringement (1) The judge had been correct in holding that a spheroid meant any spheroidal particle made up of smaller particles but had erred in interpreting spheronising agent too narrowly and in holding that the HPMC in the Cimex product was not a spheronising agent. ([30]-[34]) (2) The judge had also erred in excluding HPMC on the basis that the term spheronising agent was limited to agents which assisted in making a sphere out of something that was not a sphere and that the HPMC did not have that function. The evidence was that HPMC was suitable for four out of the five spheronisation processes known at the priority date and thus it fell within the definition in the specification that a spheronising agent was any pharmaceutically acceptable material that could be spheronised to form spheroids. ([35]-[38]) (3) The judge had erred in holding that claim 1 of 246 was not infringed because there was a significant amount of oxycodone the release of which was not controlled by the polymer layer. The claim was infringed by a tablet which had oxycodone on the outside of the coating, provided that (i) the tablet contained, in addition, oxycodone which satisfied paragraphs (a) to (c) of the claim [as set out above], and (ii) the tablet as a whole satisfied the dissolution requirements of paragraph (d). ([43]) (4) Had claim 1 of 246 stopped at the end of paragraph (c), the issue whether it covered a tablet, which conformed to the terms of paragraphs (a) to (c) but had an additional application of oxycodone on the outside of the coating, would have been relatively simple. The fact that such a tablet had an additional feature would not have prevented it infringing claim 1 of 246 because such tablet would be a controlled release dosage form, although it would also have included dosage for immediate release. ([44]) (5) Paragraph (d) of the claim, however, presented two possible difficulties for this conclusion. First, the dissolution rate would not have been entirely attributable to the oxycodone described in the preceding three paragraphs: (d) would be dissociated from (a) to (c). Secondly, it would have meant that the oxycodone salt referred to in paragraph (d) would not have been precisely the same as the oxycodone salt referred to in (a) to (c). While these points were not without force, they were not persuasive enough to restrict claim 1 of 246 as contended by the respondents. Whilst paragraph (d), on this reading, was somewhat disconnected from paragraphs (a) to (c) of the claim, such a disconnection was consistent with paragraphs (b) and (c) referring expressly to the said oxycodone salt, tying them into paragraph (a) where such a salt was mentioned, whereas paragraph (d) was tied into the opening line of the claim, by virtue of its reference to the said dosage form. Accordingly, in the case of a tablet with externally applied 4 Relying on Hallen Co v Brabantia (UK) Ltd [1991] R.P.C. 195, CA.

6 544 NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GMBH oxycodone in addition to oxycodone within the coating, the dosage form was all the oxycodone, whereas the said oxycodone salt was that within the coating. Further, there was no logical, technical or commercial reason why the dissolution rates in paragraph (d) should not have extended to all the oxycodone in the case of a tablet with external oxycodone, as well as oxycodone within the coating. ([46],[47]) H28 (6) The Cimex tablets infringed claims 1 and 4 of 246 and claim 6 of 730. ([39],[41],[48],[62],[117]) Added matter H29 (7) PPG Industries did not set up any further or more extensive rule than the basic rule that an undisclosed disclaimer was permissible as not adding matter provided it was a mere disclaimer. ([82]) Case G0001/03, PPG Industries Ohio, Inc, 8 April 2004, Enlarged Board of Appeal, EPO and Decision T-1139/00 CORDIS CORPN/Balloons for medical devices, 10 February 2005, Technical Board of Appeal, EPO considered. H30 (8) The judge had been entirely right in holding that the test for added subject matter remained that set out in the Convention and the Act; that the reason that disclaimers of accidental and deemed anticipations did not offend was that they did not add subject matter relevant to the invention; and that if a disclaimer introduced by a divisional application did not add subject matter relevant to the invention, but merely excluded subject matter from protection, then it too would not offend against the provision. ([85]) H31 (9) It was not correct to say that there was a disclosure in 730 of two classes of phindependent acrylic resins. Both in the application and the patent the ph-independence was that of the formulation, not the resin as such. The judge had been correct in holding that the skilled person would have read 730 as teaching that acrylic resins were suitable materials to use for achieving ph-independent release, but that the combined effect of the disclosure in the specification and the disclaimer was that these materials were not claimed. He would not have derived from the general disclosure any technical teaching about ph-dependent acrylic resins. ([94],[95]) H32 (10) The attack based on added matter failed. ([97]-[99]) H33 H34 H35 H36 Lack of inventive step (11) The judge s rejection of the respondents expert s evidence that oxycodone would have been included on a shortlist of known drugs as suitable alternatives for controlled release morphine was crucial. Coupled with the finding that oxycodone was not known as anything other than a co-drug and was not perceived as an alternative to morphine, the case on common general knowledge had to fail. ([104]) (12) The judge s summary of the differences between the inventive concepts of the claims in suit and Oshlack was accurate. The judge s finding that the skilled person would not have set about modifying Oshlack so as to achieve a 12-hour slow release formulation was well within the permissible range of value judgments open to him. ([101],[111]) Biogen Inc v Medeva plc [1997] R.P.C. 1, HL referred to. (13) The respondents case on appeal was rejected. It made no sense to say that a formulator would press on and develop a twelve-hour release formulation without any apparent clinical point. Motive remained relevant on any view. ([113]) (14) There was a real difference between the Hallen v Brabantia sort of case and the present. There it was obvious to put PTFE on any corkscrew and that a worthwhile advantage would be obtained. Here it was not obvious that any advantage would be Published by Oxford University Press for the Intellectual Property Office

7 [2009] R.P.C H39 H37 H38 obtained, for there was no demand for an oral, still less a 12-hour (or any) slow release, formulation of oxycodone. An unexpected advantage only failed to defeat an obviousness attack where there was a real motive to use the idea apart from that advantage. For only then would the skilled man more or less inevitably bump into the unexpected advantage and even that might not be enough to destroy a patent, for there might possibly be room in some cases for an invention by selection. ([114]-[115]) Hallen Co v Brabantia (UK) Ltd [1991] R.P.C. 195, CA distinguished. (15) The appeal on infringement was allowed and the cross-appeal on validity was dismissed. The patents were valid and infringed. ([117]) Cases referred to: A.C. Edwards Ltd v Acme Signs & Displays Ltd [1992] R.P.C. 131, CA Decision G1/93, ADVANCED SEMICONDUCTOR PRODUCTS/Limiting Feature [1995] E.P.O.R. 97, Enlarged Board of Appeal, EPO Biogen Inc v Medeva plc [1997] R.P.C. 1, HL Bonzel v Intervention (No 3) [1991] R.P.C. 553, Pat Ct Decision T-1139/00 CORDIS CORPN/Balloons for medical devices, 10 February 2005, Technical Board of Appeal, EPO Hallen Co v Brabantia (UK) Ltd [1991] R.P.C. 195, CA Merrell Dow Pharmaceuticals Inc v HN Norton & Co Ltd, unreported, 1 October 1996, Pat Ct Pozzoli SpA v BDMO SA [2007] F.S.R. 37, CA Decision G0001/03, PPG Industries Ohio, Inc 8 April 2004, Enlarged Board of Appeal, EPO Vector Corp. v Glatt Air Techniques Ltd [2008] R.P.C. 10, CA Windsurfing International Inc v Tabur Marine (Great Britain) Ltd [1985] R.P.C. 59, CA Michael Tappin instructed by Powell Gilbert LLP appeared for the appellant/defendant in each action. Michael Silverleaf, Q.C. and Piers Acland instructed by Nabarro Nathanson LLP appeared for the respondent/claimant, ratiopharm GmbH. Michael Silverleaf, Q.C. and Mark Chacksfield instructed by SJ Berwin LLP appeared for the respondent/claimant, Sandoz Ltd. JUDGMENT 1 JACOB L.J.: This is the judgment of the court to which all members have contributed. 2 We have an appeal (argued by Mr. Michael Tappin) and cross-appeal (argued by Mr. Michael Silverleaf Q.C.) from a judgment of Floyd J. of 16 December The appeals were expedited because the generic medicine parties, ratiopharm GmbH and Sandoz Limited (collectively r/s ), are about to receive a marketing authorisation. Determining a case on its full merits quickly generally means that there is no need for any provisional measures or the legal costs and time which would otherwise be devoted to them. That was so here. 3 The patentees are Napp Pharmaceutical Holdings Ltd. Napp owns the two patents in suit, Nos. EP (UK) and The judge held both patents valid but not infringed by the tablets which r/s wish to sell, a product made by Cimex AG (now Acino AG). Below it was called the Cimex product and we will do the same. 5 [2009] R.P.C. 11, Sub nom ratiopharm GmbH v Napp Pharmaceutical Holdings Ltd.

8 546 NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GMBH 4 The patents relate to controlled release formulations of a painkiller called oxycodone. Napp sells such a formulation under the trade mark OxyContin. Its importance is shown by its sales figures of the order of 32 million per annum in the UK alone. 5 OxyContin is a painkiller used as an alternative to morphine for moderate to severe pain. It has certain advantages as compared with morphine. r/s s own expert witness, Professor Bennett said, as recorded by the judge at [258], that it formed a valuable weapon in the armoury of the pain management clinician working in palliative care or otherwise. 6 Even if the decision that the Cimex product does not infringe were upheld it would be necessary to consider validity. Bifurcation of the issues of infringement and validity would serve no useful purpose. This is because Sandoz has another product in the pipeline which it is said infringes some of the claims of the 730 patent. We do not know whether it does or not: there is an infringement action pending. Divisional Patents 7 The two patents have, for practical purposes, the same text because they are divisionals. The differences lie in their respective claims, and in variations of the text consequential upon the dividing out process. For present purposes it is common ground, as it was below, that we can work mainly from the text of We should say a word about divisionals. The basic procedure for applying for a patent is laid down in Art. 75 of the EPC. This allows an inventor to apply for a patent by filing his/her application at the EPO (or a national office of an EPC member if the law of that state allows it). The relevant date for the purposes of judging patentability (the priority date ) will be the date of that application, though an earlier priority date may be claimed if the patentee has made an earlier (but within one year) application in a Paris Convention country. The application can be made to the EPO via the machinery of the Patent Co-Operation Treaty (PCT). That was used to make the application leading to the patents in suit in this case. 9 However Art.76 of the EPC allows for an elaboration of the basic procedure. It goes by the jargon divisional. Art.76 says: 76(1) A European divisional application must be filed directly with the European Patent Office at Munich or its branch at The Hague. It may be filed only in respect of subject-matter which does not extend beyond the content of the earlier application as filed; in so far as this provision is complied with, the divisional application shall be deemed to have been filed on the date of filing of the earlier application and shall have the benefit of any right to priority. 10 So what a patentee can do, having made an initial application, is to apply for a divisional patent. Provided the subject-matter of this does not extend beyond the content of the earlier application, he can get a free-standing patent for the divisional application. Because the date of filing is deemed to be that of the parent as the jargon goes, a patentee cannot extend the period of protection by applying for a divisional. 11 Sometimes the EPO will say to a patentee that he has got too much in a single application (an application can only cover a single invention or group of inventive concepts, Art.82) and he must divide out or cut down. But a divisional can be sought by a patentee even if the Office does not require it. Then the patentee s purpose is to break up his/her original application into a cluster of sister patents each having the same basic disclosure but with different claims. Much the same could be achieved by just adding these claims to the original application, but there are both real and perceived Published by Oxford University Press for the Intellectual Property Office

9 [2009] R.P.C advantages in using divisionals. Firstly each patent will stand or fall on its own merits whereas otherwise there is a risk of the complications of a partially valid patent complications with potentially different results in different Member States. Secondly a possible infringer can be sued on just the divisional(s) relevant to what he does. Thirdly if the patentee is having trouble with the examination process in respect of part of his invention, he may divide out the portion which the Office accepts is all right so that it gets early grant of that, and meanwhile he can pursue his argument with the Office in respect of the remainder. 12 It is fair to say that a clutch of divisionals is likely to make the position more difficult to assess for third parties. Here for instance there are no less than nine divisionals stemming from the original application. It is perhaps questionable as to whether the current voluntary aspect of the divisional system should continue to be permitted. That is particularly so since one of its justifications has now gone or is of less significance. A patentee need no longer fear that by having all his claimed eggs in one basket, he will lose his patent if even one egg turns out to be bad: central amendment of a granted patent is now possible pursuant to the provisions of Art.105a of the EPC (introduced by amendment pursuant to the EPC 2000). 13 One of the features of the divisional system is that each divisional must have claims which are different: the patentee cannot have the same claim in different patents. So filing a divisional involves a dividing up process: what is claimed in the divisional must be excluded from the claims of the patent from which it has been divided out. That dividing out process may, perhaps, itself add subject-matter it is said to have done so in this case. Technical Background 14 The judge describes this uncontroversially at [8] to [25]. To summarise: (i) Extended release formulations of pharmaceuticals for oral administration were well known. The tablet is designed so that the drug is released slowly over a period of time rather than immediately. The idea is to keep the level of drug in the body near constant. (ii) Various methods were used to achieve this: in particular (a) a tablet consisting of a matrix containing the drug where the matrix dissolves slowly, releasing the drug as it does so and (b) a tablet containing the drug contained in granules surrounded by a barrier the drug leaches slowly through the barrier. (iii) A particular form of the barrier type tablet involves coating non-pareil sugar beads with the drug and adding the barrier on top. There are other well-known ways of enclosing the drug. (iv) Opioid analgesics were well-known. These act on opioid receptors. They consist of the naturally occurring alkaloids (including morphine and codeine), the semi-synthetic opioids (made from naturally occurring substances they include heroin, the drug of interest in this case, oxycodone, and others), fully synthetic opioids and opioid peptides. (v) Some opioid analgesics are stronger than others. The gold standard for pain killing was morphine, but it had significant side effects, including dependence. Moreover the dose appropriate for a particular patient had to be found by a lengthy process of trial and error (called titration ), there being considerable variation from patient to patient.

10 548 NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GMBH (vi) There was a strong desire for an alternative to morphine with less side effects and easier titratability. The Patents 15 Broadly, the two patents in suit are for a 12-hour controlled release form of oxycodone. The claims alleged to be infringed by the Cimex product are for barrier-type controlled release forms. The barrier-type form specifically described is made by the use of extrusion spheronisation, one of a number of well-known way of making small spherical particles. The description also includes matrix-type forms no doubt they form the subject of one or more of the divided out patents as well as some claims of the 730 patent. 16 The judge describes the 730 patent at [35]-[49], citing a number of passages which are of significance in relation to claim construction. We will come to them at that point in our judgment. For present purposes it is important to note what the inventors say is the real point of their invention. It is effectively contained in two main passages, though there are also others cited by the judge: [21] It has now been surprisingly discovered that the presently claimed controlled release oxycodone formulations acceptably control pain over a substantially narrower, approximately four-fold (10 to 40 mg every 12 hours around theclock dosing) in approximately 90% of patients. This is in sharp contrast to the approximately eight-fold range required for approximately 90% of patients for opioid analgesics in general. [23] Morphine, which is considered to be the prototypic opioid analgesic, has been formulated into a 12 hour controlled release formulation (i.e. MS Contin R tablets commercially available from Purdue Pharma, L.P). Despite the fact that both controlled-release oxycodone and controlled release morphine administered every 12 hours around-the-clock possess qualitatively comparable clinical pharmacokinetic characteristics, the oxycodone formulations of the presently claimed invention can be used over approximately 1/2 the dosage range as compared to commercially available controlled release morphine formulations (such as MS Contin R ) to control 90% of patients with significant pain. 17 These two passages along with the other passages are saying two things: that a 12- hour slow release oxycodone is as good at killing pain as morphine and that it needs less titration compared with a slow release morphine because there is less patient-to-patient variation. Although the latter advantage was challenged and the judge was not satisfied that it is so, the former was not and as we have said the invention has in fact proved to be a valuable alternative to slow release morphine. The Claims 18 Turning to the claims of the patents it is only necessary to set out those relevant to our decision (the judge set out rather more claims in an annex). Those claims are: Claim 1 of 730 A controlled release oxycodone formulation for oral administration to human patients, comprising: Published by Oxford University Press for the Intellectual Property Office

11 [2009] R.P.C (a) oxycodone salt in an amount equivalent to 10 mg to 160 mg of the oxycodone hydrochloride salt, and (b) a controlled release dosage matrix, other than an acrylic resin matrix selected so that the formulation provides ph-independent dissolution characteristics, (c) wherein said formulation provides, at steady state after repeated administration at 12-hour intervals, a mean maximum plasma concentration of oxycodone of 6 to 240 ng/ml at 2 to 4.5 hours after administration and a mean minimum plasma concentration of oxycodone of 3 to 120 ng/ml at 10 to 14 hours after administration. Claim 6 of 730 A controlled release oxycodone formulation for administration to human patients, comprising: (a) an analgesically effective amount of spheroids comprising oxycodone or a salt thereof and a spheronising agent; (b) each spheroid being coated with a film coating which controls the release of the oxycodone or oxycodone salt at a controlled rate in an aqueous medium. Claim 7 of 730 The controlled release oxycodone formulation of claim 6, comprising: an analgesically effective amount of spheroids comprising oxycodone salt and a spheronising agent, such that the total dosage of oxycodone salt in said dosage form is from 10 to 160 mg. Claim 1 of 246 A controlled release oxycodone dosage form for oral administration to human patients, comprising: (a) oxycodone salt in an amount equivalent to 10 mg to 160 mg of the oxycodone hydrochloride salt; (b) a matrix incorporating said oxycodone salt; (c) a coating on said matrix controlling the release of said oxycodone salt; (d) wherein said dosage form has an in vitro dissolution rate, when measured by the USP Paddle Method at 100 rpm in 900 ml aqueous buffer (ph between 1.6 and 7.2) at 37 C, between 12.5% and 42.5% (by weight) oxycodone salt released after 1 hour, between 25% and 55% (by weight) oxycodone salt released after 2 hours, between 45% and 75% (by weight) oxycodone salt released after 4 hours and between 55% and 85% (by weight) oxycodone salt released after 6 hours. Claim 4 of 246 The controlled release oxycodone dosage form of any one of claims 1 to 3 comprising: (a) an analgesically effective amount of spheroids comprising oxycodone salt and a spheronising agent; (b) each spheroid being coated with a film coating which controls the release of the oxycodone salt at a controlled rate in an aqueous medium.

12 550 NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GMBH The issues on the appeal/the structure of the Cimex product 19 Because there were a surprisingly large number of issues raised on the appeal, we asked the parties to provide an agreed list of issues in advance of the hearing. This proved to be helpful. We do not set it out here, preferring to set out each issue as we come to it. 20 And also before coming to the issues of construction, to explain why they are issues, it is helpful to describe the alleged infringing product, the Cimex tablets. We borrow with gratitude the judge s description: [98] The Cimex product which is accused of infringement is a tablet which contains, embedded in excipients, small particles made up of a number of layers. A cross S. of a particle showing the layers is shown below. [99] The Core is bought in by Cimex. It is made from sugar crystals which are built on with liquid glucose and corn starch in a standard coating vessel from which they are taken and dried. The Inner Layer is made up of oxycodone hydrochloride (83 wt%), hydroxypropylmethyl cellulose (HPMC, about 8 wt%) and talc/macrogol (about 9 wt%). It is sprayed onto the cores in a fluid bed processor. [100] The Polymer Layer contains ethyl cellulose, hydroxypropylcellulose and propylene glycol. It is accepted by ratiopharm/sandoz that the Polymer Layer controls the release of the oxycodone hydrochloride in the Inner Layer. [101] The Outer Layer contains about 96 wt% oxycodone hydrochloride and about 4 wt% HPMC. About 20% of the total weight of oxycodone hydrochloride in the dosage form is in the Outer Layer. 21 In brief, therefore, 20% of the oxycodone is on the outside of the particles and 80% is within. The 20% will be released more or less immediately whereas the 80% will be slow-released through the polymer layer. By contrast (apart from the fact that the particles are made by a different process, involving building them up from non-pareil Published by Oxford University Press for the Intellectual Property Office

13 [2009] R.P.C beads) the key difference from what is specifically described in the patents is that in the latter all the oxycodone is inside a polymer layer and will be control-released. In the course of argument, the difference was likened to ordinary as opposed to chocolate coated Smarties (M&M s to some readers). CONSTRUCTION/INFRINGEMENT A. Spheroid and spheronising agent Claim 6 of 730, claim 4 of 246 The issues and the judge s conclusions 22 r/s did not admit that the Cimex tablets comprised spheroids or a spheronising agent. They pleaded that the term spheroids denoted spherical granules manufactured by spheronisation and that none of the excipients constituted a spheronising agent. The inner and outer layers contain HPMC, about 8 wt% and 4 wt% respectively. They say that the water-soluble HPMC serves as a binder and is not a spheronising agent. 23 The judge accepted that there were a number of ways in which spheroids could be made. They included the process of extrusion/spheronisation in which the active ingredient and excipients are passed through an extruder to produce cylindrical pieces of material. In that process, a spheronising agent was a material which gave the mass of material the right degree of plasticity when it was extruded to allow the second stage of the process to take place in an optimum way. 24 On the meaning of spheroid the judge concluded (at [80])... the skilled reader would have no reason to think that the term granule was being used in any particularly limited sense. He would know that a wide range of processes exist for arriving at a spheroidal particle by agglomerating smaller particles. He would have no reason to suppose that the patentee wanted to exclude any of them. 25 The judge recorded the submission on behalf of r/s that (a) the term spheronising agent referred to non-water soluble materials (such as microcrystalline cellulose) which were incorporated to give the correct degree of plasticity in an extrusion/spheronisation process; or (b) the term was limited to agents which had assisted in making a sphere out of something that was not a sphere. He concluded that he preferred r/s s submission, but concentrated on the second way in which they had put the argument, and found that the claim was using the term spheronising agent in the sense of an agent which had enabled the excipients to form into a sphere, and that it could not properly be applied to materials which were added once the sphere was formed. Otherwise the claim would extend to a whole host of materials which the skilled person would never dream of calling spheronising agents as they had played no significant role in the formation of the sphere. The appeal 26 r/s challenge, albeit faintly, the judge s conclusion that spheroid means spherical granule (the latter being an aggregated particle made up of smaller particles) made by any process. They say he should have held that spheroids mean spheroids produced by a process of spheronisation using a spheronising agent. This followed, they suggest, from their main submission (accepted by the judge) about the limited meaning of spheronising agent.

14 552 NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GMBH 27 Napp appeals against the judge s conclusion on spheronising agent for the following reasons. First, spheronising agent is widely defined in the specification ([42]). Second, HPMC is known as an agent for producing spheroids in, for example, the rotorgranulation process. Third, it performs that function in the production of the Cimex product by assisting in making the spheroid (and the core is not the spheroid of the claims). Fourth, that conclusion was not affected by the fact that it also functioned as a binder. 28 Mr. Silverleaf supports the judge s reasoning by reference to the following matters in particular. First, HPMC is water-soluble and unsuitable for use as a spheronising agent. The specification draws a clear distinction between the spheronising agent and the binder. The former is non-water soluble ([41]) whereas the latter is exemplified by reference to water-soluble polymers ([44]). Second, on Napp s approach, any additive used in a rotorgranulation process would qualify as a spheronising agent including water. Third, as regards the Cimex product, it defies common sense to characterise the process of building up a particle on a pre-existing spherical core by spray-coating as a process of spheronisation: the reason that the pellets are spherical is because they start with a spherical sugar core, not because of the HPMC or any other element of the Inner Layer. Our Conclusions 29 We set out those paragraphs of the specification which assist in this aspect of claim construction: [0041]...In particularly preferred embodiments of this aspect of the invention, the present dosage form comprises film coated spheroids containing active ingredient and a non-water soluble spheronising agent... [0042] The spheronising agent may be any pharmaceutically acceptable material that, together with the active ingredient, can be spheronised to form spheroids. Microcrystalline cellulose is preferred. [0043]...According to a preferred aspect of the present invention, the film coated spheroids contain between 70% and 99% (by wt), especially between 80% and 95% (by wt), of the spheronising agent, especially microcrystalline cellulose. [0044] In addition to the active ingredient and spheronising agent, the spheroids may also contain a binder. Suitable binders, such as low viscosity, water soluble polymers, will be well known to those skilled in the pharmaceutical art... [0051] The present solid, controlled release, oral dosage form may also be prepared, in the form of film coated spheroids, by (a) blending a mixture comprising oxycodone or a oxycodone salt and a non-water soluble spheronising agent, (b) extruding the blended mixture to give an extrudate, (c) spheronising the extrudate until spheroids are formed, and (d) coating the spheroids with a film coat. 30 The judge was clearly right about the meaning of spheroid. There is no purposive reason to limit it as suggested, and every purposive reason to say it means any spheroidal particle made up of smaller particles. They would all work to produce slow release, so why would the patentee have intended a limited meaning to his claims. The real point here is about spheronising agent. But the conclusion about spheroid is important to the argument about that. Published by Oxford University Press for the Intellectual Property Office

15 [2009] R.P.C We consider that the judge was wrong to find that HPMC as used in the Cimex product was not a spheronising agent within the meaning of the claims. 32 The judge recognised that the process of extrusion/spheronisation was only one of the ways in which a spheroid could be made. Other ways included rotor-granulation, a process which did not involve converting a cylinder into a sphere, but in which the spheroids are formed from a mass of powdered material. 33 Once the judge had reached the correct conclusion that spheroid was not limited to the process by which the product was made, it was wrong for him to interpret spheronising agent narrowly so as to accept the argument of r/s (if indeed he did) that it only referred to non-water soluble materials (such as microcrystalline cellulose) which were incorporated to give the correct degree of plasticity in an extrusion/spheronisation process. 34 It is true that the specification deals expressly only with the extrusion-spheronisation process (see [43] and [51] quoted above), and that that explains the references to the non-water soluble spheronising agent (especially microcrystalline cellulose). But as the judge correctly recognised other processes were not being excluded. 35 We also think he erred in excluding HPMC (as he did) on the alternative basis that the term was limited to agents which had assisted in making a sphere out of something that was not a sphere, and that the HPMC in the method used for the production of the Cimex product (active ingredient together with binder sprayed on to pre-manufactured sugar core) did not have that function. 36 The evidence was that a water soluble polymer such as HPMC was suitable for four out of the five spheronisation processes known at the priority date for producing spheroids for a controlled release formulation: (1) fluidised-bed granulation; (2) rotor granulation; (3) extended wet granulation; and (4) non-pareil beads (the Cimex process). 37 Accordingly HPMC fell within the definition in [42]: The spheronising agent may be any pharmaceutically acceptable material that, together with the active ingredient, can be spheronised to form spheroids In addition, even in the non-pareil beads method, where there is a pre-existing sphere, the HPMC assists in the formation of the relevant spheroids, i.e. those containing the active ingredient. We accept the submission by Mr. Tappin that the core was not the spheroid of the claims. The spheroid is the sphere which comprises oxycodone and a spheronising agent. It is true that [44] draws a distinction between a spheronising agent and a binder, but in our judgment the fact that HPMC also acts as a binder or matrix does not prevent it from being a spheronising agent as that term is used in the specification and the claims. 39 So the Cimex product satisfies these aspects of claim 6 of 730 and claim 4 of 246. B. Film coating which controls the release (claim 6 of 730) and coating on said matrix controlling the release of said oxycodone salt (claim 1 of 246) 40 The point here is whether, as r/s contend, claim 1 of 246 and claim 6 of 730 do not extend to a form of dosage in which some of the oxycodone is on the outside of the coating. This question arises because the Cimex product has an external layer on its spheroids consisting of about 20% of the drug, which is released immediately, the remaining 80% being within the coating which controls its release. 41 So far as claim 6 of 730 is concerned, the judge had little difficulty in concluding that a tablet which otherwise infringed was not taken out of the claim by the addition of oxycodone on the surface of the spheroids. We agree. There is nothing in the language

16 554 NAPP PHARMACEUTICAL HOLDINGS LTD v RATIOPHARM GMBH of claim 6 of 730 which excludes a tablet which has all the characteristics it sets out, but with the additional feature of an external application of oxycodone. There is also nothing in the teaching in the patent to suggest that such a tablet was intended to be excluded. Nor is there any commercial or technical reason for thinking that the patentee would have wanted to exclude an otherwise infringing article with such an additional feature. 42 Mr. Silverleaf contended that claim 7, which is based on claim 6, indicates that such an article could not be within the ambit of claim 6, as it could not be within the ambit of claim 7. For the same reasons as apply to our construction of claim 1 of 246, as discussed in the immediately ensuing paragraphs, we do not consider the basis for that contention is well founded. However, even if a tablet with some of the oxycodone on the exterior of the coating were outside the ambit of claim 7, we are unconvinced that this would have resulted in claim 6 being similarly restricted. Claim 7 is, clearly and familiarly, intended to be more limited than claim 6, and it would therefore be unsurprising if its wording resulted in an implied limitation in addition to, indeed as a result of, an express limitation on the terms of claim That brings us to the question of whether such a tablet infringes claim 1 of 246. The judge found this point difficult and so do we. However, we have come to the conclusion that claim 1 of 246 is infringed by a tablet which has oxycodone on the outside of the coating, provided that (i) the tablet contains, in addition, oxycodone which satisfies paragraphs (a) to (c) of claim 1, and (ii) the tablet as a whole satisfies the dissolution requirements of paragraph (d) of claim If claim 1 of 246 stopped at the end of paragraph (c), it appears to us that the issue whether it covered a tablet, which conformed to the terms of paragraphs (a) to (c) but had an additional application of oxycodone on the outside of the coating, would be relatively simple. The fact that such a tablet had an additional feature would no more prevent it infringing claim 1 of 246 than it would prevent it infringing claim 6 of 730. The tablet would be a controlled release dosage form, although it would also include dosage for immediate release. 45 Paragraph (d) of the claim, however, presents two possible difficulties for this conclusion. First, as Mr. Silverleaf pointed out, the dissolution rate would not be entirely attributable to the oxycodone described in the preceding three paragraphs: (d) would be dissociated from (a) to (c). Secondly, it would mean that the oxycodone salt referred to in paragraph (d) would not be precisely the same as the oxycodone salt referred to in (a) to (c). While these points are not without force, they do not persuade us that claim 1 of 246 should be restricted as r/s contend. 46 While it is true that paragraph (d), on this reading, is, as it were, somewhat disconnected from paragraphs (a) to (c) of the claim, such a disconnection is consistent with paragraphs (b) and (c) referring expressly to the said oxycodone salt, tying them into paragraph (a) where such a salt is mentioned, whereas paragraph (d) is tied into the opening line of the claim, by virtue of its reference to the said dosage form. Accordingly, in the case of a tablet with externally applied oxycodone in addition to oxycodone within the coating, the dosage form is all the oxycodone, whereas the said oxycodone salt is that within the coating. Further, there is no logical, technical or commercial reason why the dissolution rates in paragraph (d) should not extend to all the oxycodone in the case of a tablet with external oxycodone, as well as oxycodone within the coating. 47 The four references to oxycodone salt in (d) might well be thought to apply to the oxycodone salt referred to in (a) to (c), and, of course, they do, even where there is also oxycodone outside the coating. In a case where there is oxycodone outside the Published by Oxford University Press for the Intellectual Property Office