Hospira UK Generics (UK) Ltd (t/a Mylan) v Novartis AG

This appeal raises a short point about the entitlement of a patented invention to priority. The patent in suit, European Patent (UK) 1 296 689, belongs to Novartis AG, and relates to the use of a particular member of the bisphosphonate class of drugs for the treatment of, amongst other things, osteoporosis. The drug in question is called zoledronic acid although it may exist in hydrated forms referred to as zoledronate. I will use the term zoledronate in this judgment to cover all forms. Claim 7 (which is dependent on claim 5) claims the use of a zoledronate medicine for the treatment of osteoporosis where the medicine is adapted for intravenous administration in a unit dosage form which comprises from about 2 up to about 10mg of zoledronate, wherein the period between administrations is about once a year. The claim is in the so-called “Swiss” form, but nothing turns on that.

Claim 7 claims priority from, amongst other documents, United States Patent Application No 267689, referred to in this case as “PD2” because it is the second of two priority documents identified in the patent in suit. The issue, in essence, is whether PD2 discloses the subject matter of claim 7 of the patent.

The issue arose in the context of separate proceedings brought first by Hospira UK Limited and then by Generics (U.K.) Limited (trading as Mylan) for revocation of the patent. Both companies wish to market products falling within claim 7. After a trial, Arnold J decided that the patent in suit and another related patent were wholly invalid. His judgment is to be found at [2013] EWHC 516 (Pat). For present purposes it is enough to record that, as a consequence of the loss of priority, claim 7 was invalid. Novartis appeals only in relation to the judge’s finding that claim 7 of the patent in suit was not entitled to priority. It is agreed that claim 7 will be invalid if it is not entitled to priority, because of an intervening publication which would, in consequence, become available for an attack on validity. Equally, it is agreed that if claim 7 is entitled to priority then on the judge’s findings claim 7 will be valid. As claim 7 covers Novartis’ commercial product ACLASTA, the issue is of some commercial importance to the parties.

PD2 starts with a statement that the invention relates to bisphosphonates, in particular to the pharmaceutical use of bisphosphonates in the treatment of conditions of abnormally increased bone turnover, such as osteoporosis. After some introduction about bisphosphonates, and discussion of prior art proposals, the specification continues at page 2:

PD2 is therefore concerned with dosing interval. For a dosing interval to exceed the natural bone remodelling cycle, it would be longer than three months. The expression “conditions of abnormally increased bone turnover” is explained at page 3 as including postmenopausal osteoporosis, male osteoporosis and drug induced osteoporosis as well as a multitude of other conditions. “Intermittent administration” tells the reader nothing at this stage about the precise mode of administration, or the dosage size. A later passage bridging pages 3 and 4 gives more information as to the dosing interval: "In accordance with the present invention the bisphosphonate dosing interval is at least about 6 months, e.g. once every 180 days, or less frequently, conveniently once a year, or any interval in between, e.g. once every 7, 8, 9, 10, or 11 months. Dosing intervals of greater than once per year may be used, e.g. about once every 18 months or about once every 2 years, or even less frequently, e.g. a frequency of up to about once every 3 years or less often."

The skilled person would have recognised these increased dosing intervals as being of importance. As at June 2000, there were two bisphosphonates which had been approved for the treatment of osteoporosis, etidronate and alendronate. Etidronate was taken orally for 14 days at the beginning of a 90 day dosing cycle, for the remainder of which the patient took daily oral calcium carbonate. Alendronate was taken as 10 mg orally, daily. Alendronate had poor gastro-intestinal tolerance. In order to address this problem, patients were required to take the drug before breakfast with a full glass of water and then remain upright for 30 minutes, preferably being physically active during this time. These recommendations had a negative effect on patient compliance, especially given the prevalence of osteoporosis in older patients for whom such directions were more difficult to follow.

PD2 turns to mode of administration in the following passage on page 7:

Although there was some evidence that the skilled person might think “intra-arterial” as a mode of administration to be a mistake, there is no doubt that intravenous administration is being put forward as one preferred mode of administration. PD2 explains at page 8 that the particular mode of administration and the dosage may be selected by the physician, taking account of the individual patient’s age, size, life style and other characteristics.

As to dosage size, it is explained, also on page 8, that the dosage may depend on various factors, including the effectiveness and duration of action of the active ingredient and, importantly for present purposes, the mode of administration as well as the individual condition: i.e. the ailment that is being treated. PD2 states:

As these doses are expressed in mg/kg, the amounts given could result in very large unit doses. The document then says that “the dose mentioned above” is typically administered intermittently with a period of at least 6 months between doses, but it may be longer, e.g. once per year, once per 18 months or once every 2 years, or even longer, or any period in between.

PD2 also explains on page 8 that “the actual unit dose”, which I have called the dosage size, will depend upon the potency of the bisphosphonate and the dosing interval among other things. Specifically in relation to “more potent, recent bisphosphonates such as zoledronic acid”, PD2 says: “a unit dose of from about 1 up to about 10 mg may be used. For example … from about 1 to about 5 mg may be used for dosing once every 6 months; whereas a dose of from about 2 up to about 10 mg may be used for once a year dosing”.

I refer to this, simply for convenience, as “the 2-10 mg once a year passage”. It occurs immediately after a reference to single dose unit forms for infusion solution doses containing 0.5 to 500 mg of active ingredient, which would be suitable for, but not be limited to, intravenous infusion. The passage also contains a reminder that the unit dose used will depend upon potency of the active ingredient and dosing interval “amongst other things”. The “other things” clearly include those discussed earlier, including method of administration and condition.

PD2 contains five Examples. Example 5 is relevant. It describes a Phase 2 12 month clinical trial of zoledronate for the treatment of post-menopausal osteoporosis. Either zoledronate or placebo was administered intravenously. The dosage sizes and intervals in the non-placebo study arms included 4 mg every 12 months (i.e. once). Other regimens included 0.25, 0.5 and 1 mg every 3 months and 2 mg every 6 months. Total annual dosages therefore ranged from 1 to 4 mg per year. Patients were evaluated every three months over the year. All treatment arms demonstrated “a percent change from baseline in BMD significantly (p<0.001) greater than placebo and not dissimilar one from another.”

The document states at page 16:

The following conclusion is drawn:

Finally, the abstract, which the parties are agreed forms part of the relevant disclosure for priority purposes, says this:

Claim 7 contains, in combination, features directed to

Before the judge, the parties were agreed that “for the treatment of osteoporosis” is to be interpreted as meaning that the regime is in fact effective for the treatment of osteoporosis.

In order to understand the judge’s reasoning in relation to claim 7 it is necessary to look at other parts of his judgment dealing with earlier, wider claims. The judge first considered those claims which have no dosage range requirement and which merely require “at least about six months between [intravenous] administrations”. The judge held that the subject matter of these claims was not disclosed in PD2. At [137] he said this:

The judge gave as a further reason at [138] that PD2 did not disclose that zoledronate will be effective to treat osteoporosis when administered intravenously at intervals of 6 months or more regardless of dose. This is not an objection which arises in quite the same form in relation to claim 7, which does specify dose. However the judge pointed to expert evidence which indicated that the general disclosure in PD2 about the dosage range could not be taken to apply to the intravenous administration of zoledronic acid. At [139] the judge rejected Novartis’ reliance on the 2-10mg once a year passage, as there was nothing to link it to intravenous administration or the treatment of osteoporosis. He relied on expert evidence which he summarised as follows:

Yet further, the judge pointed out that, despite the 2-10mg once a year passage, the only teaching which the specification gave as to those dosages was in Example 5 which gives results only for a total dose of 4 mg per annum. He also considered that Example 5 was limited in its disclosure in that it only created an expectation that there would be fracture reduction in osteoporosis patients, there was not actual disclosure of efficacy in fracture reduction.

Turning to claims 5 and 7 the Judge held that there was nothing to link the dosage sizes and intervals there claimed with the other features of the claims, such as treatment of osteoporosis and intravenous administration.

Starting with the 2-10mg once a year passage, Novartis submits that there is in PD2 a clear and unambiguous disclosure of the use of 2-10mg of zoledronate once a year. Although osteoporosis is not expressly mentioned in the paragraph, the use of this regimen for the treatment of osteoporosis is part of the general teaching of the document, and indeed its focus. In addition, intravenous administration is one of the principal routes being taught, not only in the description, but in the Example. Accordingly there is a disclosure of the subject matter of claim 7.

Novartis submits that the judge was led into error by failing to read the document as a whole. Had he done so he could not have failed to see that the link between zoledronate, osteoporosis and intravenous administration runs through the entire document. In holding that the only disclosure of efficacy of intravenous administration to treat osteoporosis was in Example 5 and thus limited to one dosage size, the judge was looking at the example in isolation. The judge had also read the 210mg passage in isolation.

The Respondents ask this court to uphold the judge’s reasoning for the reasons he gave.

The relevant treaty provision and legislation is contained in Article 87 of the European Patent Convention (“the Convention”) and section 5 of the Patents Act 1977 (“the Act”). Article 87(1) provides:

Section 5(1) and (2) of the Act provide:

The language of the Convention and that of the Act are different: the Convention says that the priority document and the claim must be “for the same invention” whilst the Act says that the claim must be “supported by matter disclosed” in the priority document. The two formulations must, however, be taken to mean the same thing see, section 130(7) of the Act: see per Jacob LJ in Unilin Beheer BV v Berry Floor NV [2004] EWCA (Civ) 1021; [2005] FSR 6 at [39]. The language of the Convention was interpreted by the Enlarged Board of Appeal in G2/98 “Same Invention” [2001] OJEPO 413. The Board said:

We were referred to some examples of the ways in which the invention in the priority document and the patent claim have been held not to be “the same” but to be different. Whilst they are useful illustrations, they are not a basis for qualification of the general principle stated in G2/98. Thus, in Pharmacia Corp v Merck & Co Inc [2002] RPC 41, the patentee argued that it was entitled to claim priority to a claim for a class of compounds which differed from any class disclosed as such in the priority document. Aldous LJ accepted that once the patentees came to believe that the class of compounds claimed in claim 1 had particular activity it was possible to trace in the priority document where the compounds within claim 1 were disclosed. That reflected the extreme breadth of the disclosure in the priority document but it did not follow that it provided support for the claim 1. He held that the invention of the priority document was not the same as that disclosed in the application (see [100] to [101]).

Pharmacia was a case in which priority was lost by narrowing down from the disclosure of the priority document in a manner which could not be derived directly and unambiguously from it. It can be said to be a case of a new and narrow combination of individual features actually, but separately, disclosed in the priority document. One can also lose priority, however, by widening or generalising from the priority disclosure. Thus in Beloit Technologies Inc. and another v Valmet Paper Machinery Inc. and another [1995] RPC 705 at 733-734 the priority document disclosed a machine with two suction rolls, or one suction roll and a grooved roll in contrast with the claim which covered a case where there were no suction rolls at all. It was argued that the priority document disclosed “the generalised conception of a turning round”. Jacob J rejected that contention and denied the claim priority because only specific embodiments were disclosed: see pages 733-734.

In Unilin (cited above) the patent related to mating flooring panels. The claim in the patent required the panels to be connected in a manner “free of play”. It was argued that the claim was not entitled to priority because, although panels connected “free of play” were disclosed in the priority document, the disclosure was always in combination with other features. To claim “free of play” independently of these other features resulted in loss of priority. Jacob LJ rejected that argument at [61]:

Put in a nutshell, the problem for Novartis in seeking to establish that claim 7 is entitled to priority from PD2 is that the disclosure of PD2 is either too general or too specific. If one focuses on the disclosure about zoledronate, the “2-10 mg once a year” passage tells the skilled reader nothing about dosage range for any particular method of administration. It also does not tell the reader about the dosage range for any particular condition, such as osteoporosis. Example 5, on the other hand, is specific to intravenous administration. It teaches that 4 mg, once a year, administered intravenously to patients with post-menopausal osteoporosis is effective, but nothing about what other doses could be used at that dosage interval.

Mr Turner submitted that, reading the document as a whole, one sees a disclosure of the whole package of claim 7. He submits, correctly, that intravenous administration is at least one of the preferred methods of administration, and, again correctly, that osteoporosis is one of the highlighted target conditions. He argues, as he is compelled to, that the “2-10 mg once a year” passage is applicable to all modes of administration and all conditions. Thus the skilled reader is able to understand from the document as a whole that it is not just 4 mg which is effective intravenously for osteoporosis (see Example 5), but the range of 2-10 mg as well.

Mr Turner draws attention to the fact that other study arms in Example 5 use less zoledronate than 4 mg per year, albeit at more frequent dosing intervals. Thus 4 mg would not be understood as the only possible dose, but an example of a suitable dose within a suitable range. At least one delineation of the suitable range can be found in the “2-10 mg once a year” passage.

I am unable to accept those submissions. They depend, critically, on reading the teaching of the “2-10 mg once a year” passage as saying that, no matter how you administer zoledronate, and no matter what condition you administer it for, 2-10 mg is a suitable dosage range. To put it another way, it would be read as saying that this particular dosage range can be used independently of the condition being treated and independently of the method of administration. There is, however, an alternative way of reading it, namely that, depending on the method of administration and the condition being treated, some doses within this range may be suitable.

Like the judge, I have no doubt that the skilled person would read the passage in the latter sense. Firstly, the specification expressly makes the point, at least twice and for the second time just before this passage, that dosage is dependent on method of administration and condition. Mr Turner sought to dismiss the first of these passages as no more than “boiler plating”, but I do not think the skilled person would take that approach. Secondly, the skilled person knows from his or her common general knowledge that dosage is critically dependent on condition and method of administration. The intravenous route delivers the drug directly to the bloodstream without the loss of drug via the excretion that would occur in the case of oral administration. A lower dose would obviously apply in such cases. Thirdly, other dosage ranges are given in the patent which, on any view, could not be taken as saying that they were suitable for every condition and every means of administration. For example, the reference to a dose of 0.005 - 20 mg/kg (as per the passage in PD2 set out in [9]) would at a dose of 20mg/kg result in an enormous dose for a 75 kg man. Likewise the reference to ampoules for infusion containing up to 500 mg of active. This is not a promising context for a conclusion that the 2-10 mg once a year passage is of general application.

The judge also had evidence from the two experts whose evidence he summarised at [139] of his judgment. Professor Compston, who is the Emeritus Professor of Bone Chemistry at the University of Cambridge, School of Clinical Medicine and was until December 2012 Honorary Consultant Physician at Cambridge University Hospitals NHS Foundation Trust had said in her report that she believed the skilled team would think that the 2-10mg once a year passage was intended to account for the various different modes of administration contemplated. Some of the range would be appropriate to intravenous administration and some of it (likely to be the higher end) would be appropriate to other modes with lower bioavailability such as transdermal. We were shown some cross-examination about this, but it did not seem to me to take the matter any further. Her view as expressed in her second report accords with the natural reading of PD2.

Mr Turner suggested that the reference on Hospira and Mylan’s side to transdermal (and not oral) administration reflected the fact that oral administration would require much higher relative doses if 2-10 mg was appropriate for intravenous. This was so because of the very low percentage (a few percent) which would enter the bloodstream by oral administration. But considerations such as this do not get Mr Turner where he needs to be, i.e. that the passage discloses 2-10 mg for intravenous administration.

Professor Russell, who is Professor of Musculoskeletal Pharmacology in the Botnar Research Centre at Oxford University explained that it was fairly obvious that the 210 mg range would be the sort of range of doses that would be “in play” for intravenous administration. As the judge held, however, that is well short of a disclosure that this range was suitable for treating osteoporosis by intravenous administration.

There was certainly nothing in the expert evidence to displace the view that there is no disclosure in PD2 of using 2-10 mg zoledronate once a year by intravenous administration to treat osteoporosis. If that is the case, claim 7 is not in respect of an invention which is disclosed in PD2, and it is not entitled to its priority date.

Both the argument on this appeal, and the judgment of the learned judge, touch on the question of efficacy. I have reached a conclusion on this appeal simply on the basis of what is disclosed by the priority document. If it were possible to read the 2-10 mg once a year passage as disclosing that particular dosing regimen for intravenous administration for the treatment of osteoporosis, then I think one would conclude that the patentee was teaching that the regimen would be effective. As Mr Turner submitted, the judge’s construction of “for the treatment of osteoporosis” would be understood in the same sense in both the patent and the priority document. But without that disclosure, the argument does not get off the ground.

I consider that the judge was right to conclude that claim 7 was not entitled to priority from PD2. It follows that the patent is invalid. I would dismiss the appeal.

I agree.

I also agree.