1. Three generic drug companies, Teva, Arrow, and Generics (UK) (whom I shall collectively call "Teva") seek revocation of 2 patents, Nos. 2,118,042 ("042") and EP (UK) 0,998,292 ("292"). 042 is held by Istituto Gentili and 292 by Merck. Merck now owns Istituto Gentili and I shall collectively call the two patentees "Merck". Both patents are concerned with a pharmaceutically active compound called alendronate but differ widely in their priority dates. These are 1982 and 1997 respectively, nothing turning on the precise date or any question of priority.

2. Each patent of course requires quite separate consideration, but nonetheless there is some common background material necessary to understand them. There is a helpful Primer prepared by both parties which sets in more detail some of this material. A convenient précis of some of this is in Merck's closing submissions. I borrow from both documents with gratitude.

Biological facts and relevant medical conditions

3. Bone is made up of crystals of hydroxyapatite (calcium phosphate) that are arranged within an organic matrix, which is made primarily of the protein collagen. Bone has the ability to renew itself by a process called bone remodelling. Old bone is constantly being removed (or resorbed) by cells called osteoclasts and replaced by new bone by cells called osteoblasts. Normally, in a fully grown adult, the rate of bone resorption equals the rate of bone formation, such that the total bone mass remains constant. In certain diseases, the remodelling process is disturbed (usually by increased bone resorption). Such diseases include:

(i) Osteoporosis, where there is an imbalance between bone resorption and bone formation, in favour of the former, resulting in increased bone loss and destruction of bone architecture, leading to an increased risk of fractures. It is prevalent in post-menopausal women, in whom bone resorption is increased, in part, due to a loss of oestrogens that follow the menopause;

(ii) Paget’s disease, where, in one or more bones, the overall rate of bone remodelling is increased (i.e. both resorption and formation proceed at an increased rate). This leads to changes in shape and size of the affected bones;

(iii) Malignancy associated bone disease, where cancer cells cause an increase in the rate of bone resorption.

(iv) Osteomalacia. This is caused when the collagen matrix formed by the osteoblast (i.e. bone forming) cells fails to mineralise with calcium phosphate. Consequently, the resulting bone has decreased strength and can fracture easily.

Urolithiasis (formation of urinary stones) is another condition relevant to this case. It is not a bone disorder. It is a specific form of calcification that occurs in the urinary tract. Calcium phosphate crystals deposit on sites which are not normally calcified, such as the walls of blood vessels.

The bisphosphonates

4. Bisphosphonates, formerly called diphosphonates, are compounds characterised by two carbon-phosphorus bonds. The P-C-P bisphosphonates with which this case is concerned have the following general chemical structure:

5. The two R groups can be varied. By 1982 three bisphosphonates had been studied clinically. Their structures and shortened chemical names were as follows:

6. It will be observed that the R₁ and R₂ groups for etidronate are hydroxyl and methyl, for clodronate, two chlorines and for pamidronate a hydroxyl, and an amino attached to a C₂ alkyl. It is often described in this case as the "C_3" compound - counting all the carbons. Alendronate is similar to pamidronate, save that the alkyl chain is one C longer. It is the C₄ compound. So its structure is this:

7. Its full chemical name is 4-aminobutylidene-1-hydroxy-1,1-bisphosphonic acid. The full name is based on the central alkyl chain which is 4 carbons long (hence butyl). The carbon numbered 1 has a hydroxy (hence "1-hydroxy") and the two phosphonic groups (hence 1,1 bisphosphonic). Carbon 4 has the amine group (hence "4-amino"). Actually the term alendronate (like the others) is used not only for the acid but also for its salts.

8. Other aminohydroxybisphosphonates featuring in the case are the C₅ (which has no short name) and the C₆, called neridronate.

9. In a number of places different terminology is used. To simplify matters I have used the above-names wherever I can - for instance I have used these names in my quotations from the patents rather than those actually used.

The 042 Patent

10. I begin by going straight to the claims said to have independent validity. They are:

1. A pharmaceutical composition which contains as the active ingredient at least one [of alendronate or the C₅] or a salt thereof with an alkali metal or an organic base or a basic amino acid, together with a pharmaceutically acceptable carrier or diluent. [For convenience I have stated the two compounds rather than using the "General formula I" actually in the claim].

2. A pharmaceutical composition according claim 2 in a form for oral administration

7. A pharmaceutical composition according to any one of the preceding claims which contains as an active ingredient [alendronate].

11. The patent states that the invention:

"especially relates to pharmaceutical compositions suitable for the treatment of urolithiasis and capable of inhibiting bone reabsorbtion."

12. The specification continues by acknowledging some general background prior art, in particular pyrophosphate. This has the structure:

13. Although the specification does not go into detail it is common ground that the skilled man would have recognised here an acknowledgement of the late 1960's work of Fleisch's group in Switzerland. It is summarised in an uncontroversial manner by Professor Russell of Oxford University. He was a fair and in my view objective witness who gave expert evidence for the claimants. His summary said:

"Fleisch’s group, of which I was a member from 1964 onwards, determined that pyrophosphate binds strongly to calcium phosphate and impairs the formation of calcium phosphate crystals in vivo and in vitro. Pyrophosphate was then shown to inhibit calcification in vivo. Various types of ectopic calcification were efficiently prevented by parenteral but not the oral administration of the compound. However, there was no effect found on bone resorption. It was postulated that this was due to the rapid hydrolysis of pyrophosphate. It was this restriction on the possible uses of pyrophosphate that led to the search for analogues which would display similar physicochemical activity but which would effectively resist rapid hydrolysis."

14. All this would have been well known to the skilled man. The patent gives then goes on to acknowledge that things had moved on from pyrophosphate. It says:

"In view of the great interest connected with pyrophosphate, investigations have been carried out for the purpose of making substances with similar activity but resistant to hydrolysis. This object has been achieved partially with the synthesis of bisphosphonates, that is substances which contain the group P-C-P instead of the group P-O-P. The action of the bisphosphonates on calcium salts is similar to the action of pyrophosphate; indeed, even in low concentration they exhibit the following actions:

[details are given - essentially blocking crystallisation in vitro]

15. Again this would have been well known. The patent then says this:

"Several pharmacological and clinical studies in the scientific literature, however demonstrate that, in spite of certain analogies in activity, the several bisphosphonates used up to the present time in treatment of osteopathia exhibit some quite serious drawbacks with respect to the degree of toxicity in animals and the tolerability or the inducement of negative collateral side effects in men."

16. It then goes on to claim that the C₄ (alendronate) and the C₅ aminohydroxy bisphosphonate (Just the two compounds but identified by a "general formula I" coupled with the specification of just two alternatives for a group designated as "R") have "been now found surprisingly" to be:

"very suitable for the treatment of urolithiasis and as inhibitors of the bone reabsorption because they exhibit high activity which is not accompanied by the side effects hereinabove mentioned with respect to pyrophosphate”.

17. The only side effect of pyrophosphate "hereinabove mentioned" is rapid hydrolysis - not so much a side effect as a problem rendering it incapable of therapeutic use.

18. The patent then says:

"Several bisphosphonic acids have been described in the literature. In particular bisphosphonic acids have been described in the literature. In particular bisphosphonic acids of general formula I in which R is an unsubstituted alkyl and R' is hydroxy may be prepared by [a method is given]"

19. Now this is a bit odd: "general formula I" does not have an R' at all. It shows an R and an -OH and there are just two alternatives for R, C₄ and C₅. Plainly something has gone slightly wrong - the result of an amending down during prosecution. What this passage is saying is that the unsubstituted alkyl hydroxy compounds are known. So they were and in particular etidronate was known (C₂).

The patent then says:

"It has now been found that it is possible to prepare compounds of the general formula I in which R is an amino alkyl group and R' is hydroxy with excellent yields and with a very high degree of purity …[a reaction scheme is given]"

20. The oddity about R' continues. It does not really matter. What does matter is that the patent is here suggesting that the inventors were the first to find out how to make a hydroxybisphosphonate with an amino alkyl group and that there was something special about this process in general. In fact there seems to have been no difficulty about the chemistry. And even more significantly the patent omits to say that an amino alkyl hydroxybisphosphonate had already been made and was common general knowledge - namely the C₃, pamidronate. Actually neridronate had also been made and disclosed. Neither of these items of prior art are mentioned in the patent.

21. The patent then goes on to describe what is a general method for making the aminoalkyl hydroxybisphosphonates and then say (again oddly) that "the preferred compounds which are obtained according to the process of the invention are alendronate and the C₅ amino. There are two oddities here - the process is not claimed to be part of the invention at all, and the only compounds are these two and their salts.

22. The patent then gives examples of the preparation of the two compounds and of their characterisation. I can pass over this for nothing turns on it. It then gives details of a Toxicology Study. It was done with alendronate, the C₅ amino hydroxy bisphosphonate compound on the one hand and, "by way of comparison" the C₆ amino hydroxy bisphosphonate (neridronate) and clodronate. Nothing of significance turns on this study. Nor on the next study, an in vitro comparison of the same four compounds. It is upon the next study, the "Pharmacological tests" upon which reliance is placed as having given the world new important information. The patent says this:

"The purpose of this study is to investigate the effect of a series of novel biphosphonates on a culture of skull cells and on the bone reabsorption and the mineralization in vivo."

23. The passage said to be of most importance says this:

“It appears that the C₅ amino is the most active in inhibiting the bone reabsorption. However, there is observed a toxicity at the higher dosage. The substances alendronate and neridronate are also active on the reabsorption with a result slightly superior to clodronate. A significant difference is with respect to the mineralization because neridronate induces strong inhibition of mineralization in the dose of 10 mg of P/kg while alendronate has no effect or only a slight effect or only an effect to a very small extent.

These results show that the amino compounds with an odd number of carbon atoms are somewhat toxic but are much more active in inhibiting the bone reabsorption. The compounds with an even number of carbon atoms have an activity slightly superior to clodronate. Another significant fact is that alendronate does not induce or induces only to a very small extent the inhibition of mineralization at high dosage while neridronate exhibits high inhibition. Consequently, alendronate appears to be more suitable for use in diseases with an increased reabsorption of bone in humans.

In fact, for a long time, a bisphosphonate capable of inhibiting the growth of the crystals without affecting the bone has been the subject of research. It is concluded, therefore, that the two substances alendronate and the C₅ amino compound are destined to become medicaments capable of inhibiting the bone reabsorption.”

24. Before the claims, the patent then recites routine methods of making pharmaceutical compositions on which nothing turns.

The skilled man

25. In 1982 only a few centres around the world were researching the medical uses of bisphosphonates. The groups were Fleisch's (Fleisch was considered the "grand old man" of the subject), Sheffield (Prof. Russell was a founder), New York and Leiden (to which Merck's principal expert witness, Prof. Papapoulos is attached, having joined in 1984, though he had earlier contacts there). As far as commercial organisations were concerned, there were just two companies, Proctor & Gamble and Henkel. Until one recalls that the bisphosphonates had uses or potential uses as water softeners the names of these two companies, famous for soaps, washing powders and the like would cause a little surprise. What is perhaps surprising is that no "ordinary" pharmaceutical company seems to have shown much interest in the bisphosphonates by then. The expectations may have been that possible patent protections seemed weak so the research and risk were unlikely to be worth it.

26. Because the actual interested groups at the time did not include a research based pharmaceutical company Mr Young QC for Merck submitted, rather faintly, that a medicinal chemist would not form part of the notional research team constituting the "skilled man". That cannot be right. Anyone, but most relevantly a research based pharmaceutical company, was entitled to pick up the prior art and do what is obvious from it see per Laddie J in Brugger v Medic-Aid[1996] RPC 635 at p.653. A particular significance of that being so is that the evidence of Dr Newton, a medicinal chemist who had worked for many years in the chemical research division of Glaxo is relevant. He too I found fair and objective. A medicinal chemist's job is to identify make and test putative compounds for pharmaceutical use. Merck did not field a medicinal chemist, though with its resources it surely could have done. They did call Dr Brenner, an organic chemist whose main experience (with Merck) was in drug development rather than with the discovery of new drugs - more downstream than a medicinal chemist. He had no experience with bisphosphonates as active ingredients in pharmaceuticals. He was fair and objective, though as I indicate later, his evidence was not germane to what I have to decide.

The common general knowledge

27. The prior art is notionally to be read by the legal construct called the skilled man. He has a background of understanding which goes by the name "common general knowledge". This is knowledge which is not only known to all those in the field but is generally accepted by them as a good basis for further action, see General Tire [1972] RPC 457 at p.482. I asked the parties to formulate what they contended formed the common general knowledge for each of the patents in question in a single document so that one could readily see what was agreed and what was not. It has proved very helpful, if somewhat over-lengthy.

28. I have already set out a number of matters forming part of the common general knowledge. The following further matters were agreed, namely that:

(a) There was widespread interest in the clinical use of bisphosphonates to treat medical disorders of excess bone destruction.

(b) The oral administration of etidronate was approved for use for the treatment of Paget’s disease in the UK and elsewhere.

(c) The oral administration of etidronate, clodronate and pamidronate had been shown to be clinically effective in treating several medical disorders of excess bone destruction.

(d) The 3 bisphosphonates which had been shown to be the clinically effective in treating medical disorders of excess bone destruction had different efficacies and in particular, that etidronate had a relatively narrow therapeutic window.

(f) The existing non-bisphosphonate treatments for medical disorders of excess bone destruction did not satisfy all clinical needs and there was room for a better therapy.

(g) There were established tests which could be used for assessing the potential of a new therapeutic agent for use in the treatment of medical disorders of excess bone destruction.

29. Certain matters were not agreed or agreed only with some qualification. For instance it was not agreed as to how readily the principal established test, the "Schenk test" could be carried out. Prof. Russell said a technician could investigate 10 or 15 compounds in a fortnight. Prof. Papapoulos said it was more difficult and put forward a figure of 2 or 3. Prof. Russell described the test as it was done in practice by Schenk's team. It was quick and easy. Others may not have been as fast. On the other hand I have to say that I found Prof. Papapoulos on this and other subjects rather rigid and a little extreme. Here, for instance, he said his "post-doc fellows are complaining this is an impossible task." The Schenk test was obviously not impossible - far from it. Again Prof. Papapoulous suggested that in 1982 the state of the art was such that it was not possible to form any conclusion about the relationship between the structure of a bisphosphonate and its activity - even in the case of closely analogous structures. Another instance was his attachment of great significance to a paper by Benedict. This was published shortly before the priority date but it is not shown that Benedict was common general knowledge - indeed his paper has never received much accolade or acceptance.

30. Most of the unagreed points of common general knowledge consist of different degrees of emphasis and do not really matter. I do not propose to make findings on them. I can now turn to the attacks upon validity.

Lack of Novelty by reason of EP application No. 0 039 033A ("Blum")

31. There is no dispute that the applicable principles are those stated in General Tire [1972] RPC 457 at p.485. I do not need to repeat them yet again and turn to the citation. Blum is a process patent. It discloses a better way of making the C₄, C₅ and C₆ hydroxyaminobisphosphonic acids - just three compounds of which the first is alendronate. There is no dispute that the process described is enabling. Blum describes the uses of the products to be made by his process in the following short paragraphs:

“The amino alkanediphosphonic acid described excels by a high complexing power towards multivalent metal ions, particularly towards alkaline-earth ions and heavy metals, like iron and copper. Therefore they can be used especially for water-softening processes. Therein, it is not necessary to work with stoichiometric quantities, but by using quantities lower than stoichiometric quantities the precipitation of calcite can be considerably delayed.

Owing to their properties they are also suitable for the production of cosmetic and pharmaceutical preparations.”

32. Teva submit that this could not be clearer: Blum is specifically proposing the use of all three compounds for the production of pharmaceutical preparations. That must mean use as an active ingredient in such a preparation.

33. I think that must be right. A skilled man, aware as he would be of the C₃ compound (pamidronate) would naturally read this as a proposal to use the C₄-C₆ compounds as the active ingredient in a pharmaceutical. The point is straightforward and incapable of elaboration. However, I must consider Merck’s answers:

(a) That Blum’s comment is so terse that it would not be taken as having any meaning. That will not do. Brevity is not equivalent to meaninglessness.

(b) Next it is suggested that this would not be read by the skilled man as having any application to a potential agent in the treatment of bone disease. Professor Papapoulos so indicated in his evidence. I find this surprising. The skilled man would know about etidronate cloridronate and pamidronate and would read Blum with those compounds well in mind. Here was Blum proposing the C₄-C₆ compounds. He would read the document as proposing the three compounds for use as the active ingredient in a pharmaceutical. The idea of some other use in a pharmaceutical would not cross his mind.

Professor Papapoulos himself had no knowledge of the use of bisphosphonates other than as an active ingredient in pharmaceutical preparations. He said that he would not understand what Blum was talking about and would go to a chemist - an answer I found surprising. Well, if he went to a medicinal chemist there can be no doubt as to the answer he would have got. The medicinal chemist would undoubtedly have seen this proposal as a logical extension by way of close analogy with pamidronate. And for the reasons indicated I think a medicinal chemist is properly within the team.

(c) In any event even if Professor Papapoulos’ view were right, and that no use for bone disease would be seen, he accepted that he would have understood Blum to mean use in the treatment of ectopic calcification. That too is a pharmaceutical use as an active ingredient. It is said that those uses were of historical interest only. But that is irrelevant to any question of anticipation.

(d) Next, Merck suggest that the disclosure would be read as proposing the use of alendronate in a pharmaceutical preparation but not as an active ingredient. For that purpose they called Dr Brenner. For the reasons I have indicated he would not form part of the relevant skilled team. He indicated that in his view Blum would be pointing to the use of amino bisphosphonates for the same uses as other chelating agents had been used in pharmaceutical preparations. The sort of thing one had in mind would be stabilizers of drugs and colour additives. But he could only consider that if he did not know that the bisphosphonates were potentially pharmaceutically active. You would never consider using a pharmaceutically active compound in a pharmaceutical preparation as a stabilizer.

(e) Finally, there was an argument espoused by Professor Papapoulos of a wholly illegitimate nature based on other Blum patents belonging to Henkel. In the late 1970s Henkel had filed a couple of American patents relating to other bisphosphonates. In these Henkel had been explicit about the possibility of use as active ingredients in pharmaceutical preparations. The argument is that because Blum is not so explicit in the cited patent, he cannot have meant use as an active ingredient. The argument only had to be stated to be defeated. Moreover, it is illegitimate. There is no reason why the later Blum patents would be read with his earlier patents in mind.

34. None of these arguments persuades me that claim 1 is not anticipated. It is true that claim 1 also requires a pharmaceutically acceptable carrier or diluent. The use of these with an active ingredient is a plain consequence of the teaching to use alendronate in the production of a pharmaceutical.

35. Merck also defend claims 2 and 7 when dependent upon claim 2. Claim 2 is for a pharmaceutical composition for oral administration. Blum is not explicit about this. No formal case of anticipation is advanced, though, if it had been there would have been a lot to be said for it – see my judgment in Synthon’s Patent[2002] EWHC 2573 (Pat). The same goes for claim 7. In particular there is nothing in the nature of a selection invention by way of picking the C₄ compound. Claim 6 is to the C₅ compound - the patent does not teach a difference between them.

Obviousness over Blum

36. I begin with claim 1. Its inventive concept is the use of, amongst other things, alendronic acid or a salt thereof as the active ingredient in the pharmaceutical composition. What then is the difference between Blum and that inventive concept? If Blum is not an anticipation, the only difference is the idea of using these substances as an active ingredient.

37. The skilled man would know that pamidronate (and for that matter etidronate and clodronate) were clinically proven to be effective. In short, the skilled man would know that C₃ and C₆ compounds worked clinically. Seeing Blum would immediately put him in mind of the use of the C₄ and C₅ also.

38. Of course he would not know whether the C₄ and C₅ were better or worse than the C₃ and C₆. He would know that C₃ had proved to be successful. He may know that the C₆ was less successful, although still active (I am not actually satisfied that much was widely known about C₆₎. On the basis of this Merck suggest that the skilled man would not bother with trying the C₄ or C₅, that he would assume there is a linear fall in efficacy between C₃ and C₆ and not bother to try the C₄ and C₅. That is plain bad science for the reasons explained by Professor Russell: you cannot just draw a straight-line graph between the two and be logical. Moreover, no one knew whether or not pamidronate was the best. It was merely the best that had been found so far.

39. I conclude that the skilled man would certainly put the C₄ and C₅ compounds (and probably the C₂) high up on a testing programme and that he would want to do those tests. Dr Newton’s evidence entirely corroborates that. Claim 1 is an obvious consequence of that.

40. So also are claims 2 and 7 - it is impossible to see anything inventive about the idea of oral administration (claim 2) or that the particular compound should be alendronate (claim 7) or both together.

41. As I have said I am not satisfied that it would be part of the common general knowledge of the skilled man who read Blum that the C₆ was less active than the C₃. That depends upon his knowing about a paper called Gennari. It was delivered at conference in Switzerland in 1981. Although many interested in the field (including Prof. Russell but not Prof. Papapoulos) attended, it by no means follows that all accepted the paper. It was never actually put to Professor Russell that the teachings of Gennari were accepted by those in the field, either as a basis for further action or at all. And Prof. Papapoulos did not include the existence and properties of neridronate in his expert report summary of the common general knowledge.

Obviousness over Kabachnik (1978)

42. The disclosure here is very similar to that of Blum. It is of 5, rather than 3 compounds, one of which is alendronic acid. Kabachnik discloses their complexing ability but he says no more about their uses. What he does is to suggest that compounds of this general class have "various uses" and cross-refers to a paper by Vel'tishchev. That paper is about the clinical use of phosphonic compounds. It refers for example to the use of etidronate for Paget's disease and the fact that bone resorption was reduced by that compound.

43. The cross-reference is such that it can legitimately be made for the purposes of obviousness. Kabachnik is saying "for the uses of my compounds, see Vel'tishchev." The footnote is not just one of many which might or might not be followed up - the sort of thing lawyers and scientists put in to give their writings an air of scholarship. But actually the point does not matter. For in practice what is said in Vel'tishchev had become common general knowledge by 1982 so the skilled reader would not need the cross-reference. He would know that the bisphosphonates had clinical use. And here was Kabachnik telling you how to make 5 more - including alendronic acid.

44. So the upshot is that Kabachnik is saying his 5 new compounds have clinical uses. Given that, it is not possible to see what invention can lie in the concept of claim 1 - a pharmaceutical composition containing alendronate. Nor can I see anything inventive about claims 2 or 7.

45. Actually the position is that the case on Kabachnik is much the same as on Blum, so it is not surprising I reach the same result.

Obviousness over Fleisch (1981)

46. This paper, published only shortly before the priority date, is a general review entitled "Diphosphonates: History and Mechanisms of Action." Because it is by the "grand old man" of the subject particular attention is likely to have been paid to it. The abstract says what would have been common general knowledge:

"Many diphosphonates have been synthesized and tested and some relationship of their structure to the spectrum of biological effects have been observed. These analogues have similar properties to pyrophosphate, but unlike pyrophosphate they are resistant to enzymatic degradation. Their experimental properties have led to their clinical development as bone scanning agents and in the treatment of disorders of ectopic mineralization and increased bone resorption."

47. Fleisch points out that:

"The most effective compounds are characterised by the P-C-P bond, structures containing C-P or P-C-C-P bonds being less effective or ineffective."

48. He particularly commends the diphosphonates under the heading "Effect on bone resorption":

"In contrast to pyrophosphate, diphosphonates are extremely active in inhibiting bone resorption, …… Of all the compounds tested [clodronate] is the most potent."

49. I now come to the passage relied upon by Teva:

“A great number of diphosphonates have been investigated for their inhibitory effect of bone resorption. It appears that increasing the chain length of the C backbone

increases activity until a length of about 9 carbon atoms is reached. Adding a hydroxyl group at position 1 also increases the effect. The amino-derivatives such as [pamidronate]

are also very active. The relative activity of the diphosphonates tested is as follows: [pamidronate] > long chain 1-hydroxy diphosphonates > [clodronate] > [etidronate].”

50. The bits in bold are particularly relied upon. The article concludes with this encouragement for research:

"It would not be surprising if the first compounds tested are by no means optimal and that a further exploration of other types of disphosphonates could lead to a fruitful future development of these compounds."

51. It is clear from these passages that Fleisch was prepared to make some predictions between structure and biological activity. Dr Newton put it this way in re-examination:

Q. Could you just indicate how many activity relationships you see?

A. There is a clear activity relationship in terms of the carbon atom's chain length. There is a clear structure activity statement in terms of substitution with a hydroxyl, and then there is a further statement which is telling you that the amino compounds are extremely active as well; so there are three structure activity relationships referred to.

52. That is the heart of the obviousness case. I analyse it in Windsurfing terms, starting with the difference between Fleisch and the inventive concept of claim 1. It is this: Fleisch teaches that increasing the length of side chain (up to C₉) increases activity. He teaches that the amino group increases activity (pamidronate was his best) and he teaches that the -OH group is good for activity. What Fleisch does not actually say is take the pamidronate structure (thus keeping the beneficial features of the amino and -OH groups) and try making the chain length longer (i.e. go to alendronate, the C₅ compound and so on. It seems to me that that if the skilled man is prepared to accept Fleisch's teaching about chain length, it requires no imagination whatever to conclude that increasing the chain length will probably give rise to a more active compound than pamidronate. And so it did. Actually it was a lot more active and to that extent a surprise - one not however foreshadowed in the patent.

53. There is also this: There are not a lot of experiments to be done to test, say, the C_4-8 compounds. And no other line of inquiry is suggested by the Fleisch paper. All that is required is some laboratory work (not suggested to be onerous or difficult) to make the compounds and some biological testing. The dispute about how quickly and readily one might carry out the Schenk test might have come in here, but I do not think it matters given that on any view only a few compounds called for investigation.

54. The patent itself does not help rebut the obviousness case. At one point I thought it might, that the passage said to be of most importance (quoted above) was teaching something really new. But the difficulty is this: that the patent is written as if pamidronate and the whole idea of an amino group were completely unknown. The patent only acknowledges pyrophosphate as the prior art. It proceeds as if an amino substitution were a new idea altogether. Both the Fleisch paper and the common general knowledge of the skilled man extend well beyond what is acknowledged in the patent. So I do not think the passage helps.

55. How then is the obviousness case sought to be defeated? There are a number of suggested answers. The first is simply that everything about the bisphosphonates was so uncertain that no reasonable prediction about anything could be made. That involves the unimaginative skilled man actually disbelieving Fleisch. I can find no material to justify that.

56. An attempt to build such an argument was made. Professor Papapoulos said:

"…because the mechanism of action of bisphosphonates in inhibiting bone resorption was unknown, a rational search for better compounds was not possible."

57. But that is not the point. It does not really matter why the substances worked. Dr Newton gave unchallenged evidence that:

"Structure-activity relationships refer to changes in the activity of the drug as its structure is changed. If these follow some logical pattern then a rational change of structure may be determined even though the mechanism of action of the compound is not understood."

58. Dr Newton followed that with examples. Here, Fleisch had identified the three elements which did good: the -OH group, the amino group and the effect of increasing chain length. The logical pattern was laid out.

59. Next it was suggested that there might be problems over testing. But I am quite satisfied that sufficient testing methods existed by 1982. The minor dispute about how rapidly they could be done seems to me to be irrelevant considering how few tests would be involved.

60. Another line was to rely upon papers published after the priority date, such as one by Shinoda. This makes no sense in principle. I suppose that in theory a later paper, if it became part of the common general knowledge, might render unobvious for the future that which previously had been obvious. But what matters is what is obvious at the time.

61. Yet another line was to rely on the Benedict paper. As I have said this was not common general knowledge. That rules it out anyway. And in any event I am not satisfied that it points away from alendronate. It does not suggest that you need more than C₅ for bone resorption.

62. Finally it was suggested that the skilled man would know about the C₆ compound neridronate and that it was less active than pamidronate (C₃). So, the argument runs, you would expect the C₄ compound to be less active than the C₃ and hence not bother with it. However it was not proved that neridronate was common general knowledge so the argument breaks down at the outset. And even if it were, for the reasons which Prof. Russell gave, it would be wrong to assume a straight line fall in activity from the C₃ to the C₆ compound. There might be a rise or fall - and anyway it would certainly be worth finding out.

63. I conclude that alendronate was an obvious compound to try following the Fleisch article. There would be a high expectation of biological activity and a very good chance of increased activity over pamidronate. No one could say it was a sure-fire winner. Things could go wrong, for instance unacceptable side effects. But it was well worth finding out. And no other avenue of investigation readily suggested itself.

64. As to claims 2 and 7, given that alendronate was worth trying, the features of these would follow as a matter of course.

The 292 Patent

65. The priority date of this is July 1997. By then, of course, knowledge had moved on a lot. The following matters had become common general knowledge, as appears from the virtually undisputed evidence of Dr Compston who gave evidence for Teva:

(1) Alendronate had come onto the market in 1995, approved in two oral dosage forms: 10mg daily for osteoporosis and 40mg daily for Paget’s disease.

(2) The Paget’s treatment was only for six months, whereas the osteoporosis treatment was effectively for life.

(3) There was a side effect problem – that alendronate could affect the gastrointestinal (GI) tract. The initial recommendations for administration were stringent: that it should be taken at least 30 minutes before the first food or drink of the day with a full glass of water and that the patient should not lie down for 30 minutes after administration.

(4) Compliance with the conditions had some problems –not only would some elderly patients find them inherently difficult but also the very strict rules of the regime would not always be obeyed.

(5) Marketing approval for the use of alendronate for osteoporosis was based on a number of clinical trials which would have been well known to bone disease clinicians. From these clinicians would have known that alendronate was a potent resorption inhibitor and had good anti-fracture efficacy.

(6) Post-marketing surveillance showed that in rare cases only there were severe upper GI adverse events. Most of these were due to non-compliance – as Merck’s Dr Yates testified. It was not known whether poor compliance was the sole cause of the problem.

(7) Although Paget’s disease is much rarer than osteoporosis, a substantial number of patients had taken alendronate in the 40mg daily dose for the 6-month period required for that disease. There was no evidence that associated GI problems were any greater with that dose as compared with the 10mg dose for osteoporosis.

(8) In early 1996, because of the compliance problem, Merck made the instructions for administration more explicit – contraindicating for patients unable to comply with the 30-minute rule and those with certain oesophageal abnormalities.

(9) This was expected (and did in the event) improve matters, but (as was also expected) there would continue to be a compliance problem.

(10) Bisphosphonates had low oral bioavailabilty – which was reduced further if taken with food (hence the regime)

(11) However once absorbed they had a long duration of action.

(12) The size of their effect mainly depended on the total amount administered over time, not on the frequency of administration. Daily treatment was not critical to their effect on bone.

66. Merck seek an amendment of the patent which is no longer opposed. It is common ground that the patent stands or falls on its proposed amended claim 1. This reads:

“Use of alendronic acid, or a pharmaceutically acceptable salt thereof, for the manufacture of a medicament for inhibiting bone resorption for treating osteoporosis in a human in need thereof wherein such medicament is adapted for administration in a unit dosage form which comprises about 70mg of alendronic acid or a pharmaceutically acceptable salt thereof, on an alendronic acid active weight basis, according to a continuous schedule having a dosing interval of once weekly.”

67. The key idea is thus simple: give the osteoporosis patients 70mg once a week rather than the known 10mg a day. The justification for the idea put forward in the patent is that this produces fewer adverse GI events than the 10mg/daily regime – and, obviously, is more convenient for patients who only have to fast and stay upright for 30 minutes once a week.

68. The objective justification for the idea is based on some limited experiments with dogs reported in the patent. Dr Compston criticized these on the basis that two variables were changed at the same time so one could not properly identify which variable was responsible for what effect. This was not wholly answered by Dr Yates who gave evidence for Merck – his answers themselves depended on the unknown extent to which the GI tract recovered from any adverse effect. Clearly the information in the patent would not be enough to convince a scientist that the proposed regime had a safety advantage – Merck themselves did not know whether it would work until it was tried. Even now there is little data (and no published data) to prove this. What is clear is that the regime is a lot easier for patients and that correspondingly there is likely to be better compliance. That in itself will produce less GI problems – but whether or not there are less GI problems with properly complying patients remains moot.

Method of treatment?

69. I now turn to the objections to validity. The first of these is that the claim is to “a method of treatment of the human body by therapy” and hence deemed to be incapable of industrial application (see s.4(2) of the Patents Act 1977, implementing Art. 52(4) of the EPC). For this purpose I assume the invention is novel and non-obvious. Before considering the question further it is necessary to construe the claim. In particular is it in substance merely to a 70mg dosage form, for example an oral dose containing about 70mg of alendronate?

70. I think the answer to this is clearly that the claim is wider. The patent does not teach that a particular dosage unit is new. It teaches that a particular dosing regime is new. How that regime is achieved is not material to the inventive concept of the claim. You can take 7 10mg pills or two 35mg pills or a liquid dose measured out from a bottle. Very likely the claim covers taking two of the known and available 40mg pills too - for on a purposive construction the claim cannot be limited to a mathematically precise 70mg dose. At one point I thought the claim might be to a method of preparation of a single dose of 70mg – single pills or separately prepared liquid doses (e.g. in sachets). But this cannot be so – the patent specifically contemplates 35mg tablets which are said to be useful for administration in accordance with the invention.

71. So the patent is for the preparation of a medicament which can be used in a weekly 70mg dosing regime. Is it to method of medical treatment? But for the decision in Bristol-Myers Squibb v Baker Norton[2001] RPC 1 I would have held not, despite the width of the claim. The monopoly covers the preparation of the dose to be administered but not its actual administration. It is true that that might catch a doctor or nurse who measured out the dose prior to administration, but that would be an act of preparation not administration. However the principle decided by Bristol-Myers binds me.

72. What was that principle? Well, before coming to the judgments, it is helpful to remember what the patent was for, namely the use of taxol in the preparation of a medicament for administration to patients on a 24 hour basis. That is virtually the same as here, the only difference being that in that case the dose was patient specific, i.e. the dose had to be tailored for the individual patient. In that case, as in this, the pharmaceutical itself was not only known, but also its administration for the particular condition was known.

73. The Court of Appeal held the taxol claim was to a method of treatment. Instead of simply considering whether the monopoly covered the actual treatment, it linked the question to novelty. Buxton LJ put it this way:

“The novelty of the second medical use, on which its patentability rests, must therefore be found in applications that are new in the terms used in [two cited cases]. The novelty cannot lie in the method of use, but in the new therapeutic purpose for which the substance is used.”

74. So, in this case, the therapeutic use was old, it being the treatment of osteoporosis. What is said to confer novelty is the dosing regime. The words of Buxton LJ apply to that just as they did in taxol:

“This is not a case of a second or other medical use. It is a case of mere discovery about an old use”

75. The discovery in that case was that 3-hour infusion produced less neutropenia (an undesirable side effect) than a 24-hour infusion. The discovery here (assuming the method of treatment is novel and non-obvious) is that a 70mg/weekly regime produces less GI side effect than a 10mg./ daily regime.

76. Mr Thorley adapted the words of Holman J (at para. 111) in Bristol-Myers to this case, substituting alendronate for taxol, osteoporosis for ovarian cancer, adverse GI effects for neutropenia and the new and old alendronate dosing regimes for the new and old taxol dosing regimes. They read as follows:

“In the present case, however, the drug alendronate is exactly the same; the method of administration, orally, is exactly the same; and the therapeutic application or purpose, namely the attempt to treat osteoporosis is exactly the same. The only difference is the discovery that if the drug is administered in a unit dosage form of 70mg once weekly rather than 10mg once daily an undesirable side-effect, adverse GI effects, is less than it otherwise would be, whilst the therapeutic effect remains. No previously unrecognized advantageous properties in the chemical compound have been discovered …All that has been discovered … is that if the compound is administered once a week rather than daily, one of its disadvantageous side effects will be less than it otherwise would be.”

77. Mr Young found no answer to that. The best he could do was to try to rely upon what was said by Aldous LJ at paragraph 63:

“the form of claim 1 does not disguise its effect. The invention was the discovery that by changing the treatment from a 24-hour infusion to a 3-hour infusion a similar effect was obtained with less neutropenia. This was a discovery that a change in the method of treatment provided the result. The claim is an unsuccessful attempt to monopolise the new method of treatment by drafting along the lines of the Swiss-type claim. When analysed it is directed step-by-step to the treatment. The premedication is chosen by the doctor, and administered prior to the taxol according to the directions of the doctor. The amount of taxol is selected by the doctor, as is the time of administration. The actual medicament that is said to be suitable for treatment is produced in the patient under the supervision of the medical team. It is not part of a manufacture.”

78. I agree that this passage is rather narrower than that of Buxton LJ and Holman J, for it focuses on what the medical team is doing rather than asking whether what confers novelty is a method of treatment. But do not see that it helps Mr Young for two reasons. Firstly I have to follow the majority view, which did not depend on what was actually being done by the medical team. Secondly, on my construction of the claim there is a similar inroad into what will be done by medical teams –measuring out and administering a dose is just the sort of thing they do. And whether there is infringement would depend on what administration instructions were written on the doctor’s prescription.

79. Mr Young sought to put his case another way. He suggested that this case was different from the taxol case in that there the dosage form was old whereas here Merck were proposing not only a new regime but a new dosage form for it. The latter remains patentable, saved by s.4(3) which provides that a substance or composition may be treated as capable of industrial application if it is invented for use in a method of treatment. The fault with that argument lies in the width of claim. It covers the use of old dosage forms (7 x 10mg pills, a liquid measure from a bottle and probably 2 x 40mg) as well as a single unit 70mg dose. And in any event the majority view in the taxol case would mean even a single dose claim would be invalid, depending as it would, for novelty and inventiveness on the method of treatment.

80. I conclude that the claim is in substance to a method of treatment of the human body by therapy. I do so with regret. For patents are provided to encourage research. If new and non-obvious improved methods of administration of known drugs for known diseases are not patentable in principle – even with a Swiss form claim, then there will be less of a research incentive to find such methods. Giving the exception a very narrow scope, so that any preparation used in such a method is protectable if only by the artificial construct of a Swiss form claim, would be a research incentive. But I must follow the current state of interpretation of the exception in Bristol-Myers.

Lack of Noveltyby reason of the known 10 and 40mg pills

81. This must follow. The only difference between these and the claim lies in the method of administration - what is on the prescription. Of course it is more convenient to take a single 70mg pill than 7x10 or to measure out a liquid dose but that is not the point. In Merck's closing skeleton argument it is submitted that a once weekly administration requires a different dosage form to be manufactured. But it simply does not. It requires something different to be written on the bottle.

The witnesses on obviousness

82. Teva relied upon the evidence of Dr Juliet Compston, Reader in Metabolic Bone Diseases, Cambridge and an Honorary Consultant Physician at Addenbrooke's Hospital. Merck relied upon Prof. Papapoulos and also upon Dr Yates, who was their Vice President, US Medical and Scientific Affairs. Dr Yates, whose evidence I found fair, was essentially a witness of fact, testifying to the internal workings of Merck when the 70mg per week dosage regime was under consideration and development. He was a fair witness, though in the end the most significant material was Merck's own contemporaneous documents. Dr Compston was particularly impressive and fair.

Obviousness over Lunar News 1 and 2

83. The Lunar Corp., a manufacturer of bone density measuring equipment, published Lunar News three times a year. It was widely circulated amongst possible users of the equipment (for instance Dr Compston was on the circulation list). It was not a peer reviewed scientific paper - more in the nature of a chatty discussion of problems. That is not to say it was not widely read and taken reasonably seriously. For some reason Merck appear to have had a bit of falling out with Lunar over some of its comments - a matter which was eventually resolved by an agreement by Lunar to send proofs to Dr Yates for comments. The fact that what Lunar News said was taken seriously at Merck shows that it was not regarded as an inconsequential publication.

84. The April 1996 edition of Lunar News ("Lunar News I") said this when considering the use of alendronate for osteoporosis:

“Alendronate must be taken, after an overnight fast, 30-60 minutes before breakfast. Subjects should remain seated or standing; a very small group of patients have reported some upper gastrointestinal distress if this is not done. This regime may be difficult for the elderly to maintain chronically. An intermittent treatment program (for example, once per week, or one week every three months) with higher oral dosing, needs to be tested.”

85. The concept of claim 1 is weekly treatment at a dose of 70mg. So the difference between the disclosure and the concept is merely the identification of the dosage. Yet the known dosage was 10mg daily. 70mg per week is the same amount - and it was common general knowledge that what mattered was the total dose over time. Dr Compston said 70mg per week was "the most obvious dose to be chosen." So far as I can see she was not challenged on that. And Prof. Papapoulos, in a different context (whether there was a difference between 70 and 80mg/week) said this:

"I know that 70 represents exactly the sum of seven daily doses. It is exactly that. I know I am giving the same bioavailable amount of bisphosphonate which has been proven unequivocally to have anti-fracture efficacy, and I know where I am."

86. I prefer that part of his evidence to that where he simply said he would not chose any dosage because he would not implement Lunar News 1 at all.

87. It seems to me that a 70mg/weekly dosage is conceptually virtually spelt out by Lunar News 1. As a concept it is obvious.

88. Merck's attempt to deal with this powerful piece of prior art, (and indeed Lunar News 2) was to say that no one would follow the suggestion up - it was too dangerous. The general theme was this: there is a risk of adverse GI effects with 10mg daily. So giving a 7-fold dose, albeit only weekly, would severely increase that risk. To support this Dr Compston was taken to a number of papers. The upshot was put to her thus:

Q. You agree that your views are different from these other writers that I have referred to?

A. I think it is the difference between perception and evidence. I agree there was a perception amongst some people that there was definitely a dose response in the upper GI adverse events. What I do not accept is that there was actual evidence to support that.

89. I think she was right about that - the cited papers do not really contain concrete data supporting the, perhaps slightly intuitive, notion that the greater the dose the greater the damage. Perhaps the best way of assessing the position at the time is to be found in Merck's contemporaneous documents. To test whether there were any problems with the larger but less frequent dosage some dog studies were done. These are in the patent. To get permission to market the product the FDA had to be convinced. That meant looking at all the data on humans, including the papers put to Dr Compston. This is what was said under the heading "Evidence for Safety":

Paget's disease treated with 80mg/day for 3 to 6 months in 42 patients with good tolerability (…).

Paget's disease 40mg dose is generally well tolerated in Paget's clinical trials - although modest excess of discontinuations due to nonserious UGI AEs in postmenopausal osteoporosis population.

Few reports of UGI AEs from marketed use of 40mg.

90. In short it was Merck's view that the published literature did not support an unacceptable danger if 70mg/weekly was administered. It is true they did the dog studies too. That was only prudent before proceeding to try it on people. One could only ultimately find out if there was an adverse reaction by doing that. Interestingly Merck did not know that when they filed the patent. And as I have said it is still not known whether the new regime inherently has less side effects or is simply better because compliance is better.

91. I should add that Prof. Papapoulos was dismissive of the safety data for 40mg/daily for Paget's disease. The suggestion was that patients would be under tighter medical control with corresponding better compliance. But that does not help - for better compliance was one of the obvious consequences of a once weekly regime. And it is noteworthy that Merck themselves thought that the Paget's experience was evidence of safety.

92. It follows that the "fear" point fails - and with it the defence to the obviousness claim.

93. The same is true of Lunar News 2 (July 1996). This said:

“Some US physicians are reluctant to treat because of: (a) side effects, (b) difficulty of dosing, and (c) high costs ($700/year). First, Merck recently sent a letter to physicians warning of esophagitis. Some physicians report that 5 to 15% of patients experience gastric and/or esophageal distress, but most have seen no side effects. Serious side-effects of ulceration and stricture appear rare [Maconi et al]. Second, some patients also stop alendronate because of the dosing difficulty. The limited bioavailability of alendronate (0.8%) requires that it be taken on an empty stomach upon awakening with a full glass of water (not tea, coffee or juice), and the patient must remain upright for 30 to 60 minutes [Gertz et al]. A few elderly women can tolerate this regime for only a week or two…….

……The difficulties with oral bisphosphonates may favour their episodic (once/week), or cyclical (one week each month) administration. Even oral alendronate potentially could be given in a 40 or 80 mg dose once/week to avoid dosing problems and reduce costs.”

94. This is more specific about the dosage. Probably it mentions 40 or 80 rather than 70mg/ weekly because there was a known 40mg pill. It is even closer to the inventive concept of the claim. The "fear defence" to this fails for the same reason as for Lunar News 1.

Conclusion

95. I accordingly hold both patents invalid. I do so with some regret. Merck have only had a few years' exclusive exploitation of alendronate. They must surely have had to make a very considerable investment and incurred considerable risk in bringing it to market. And mankind is better off as a result. But the patent system does not confer monopolies on those who develop obvious or old products, even if they have never been exploited. A workable system for that might be a good idea, particularly in the field of medicine and analogous fields.