England and Wales High Court (Patents Court) Decisions

1. These proceedings started life as actions for a declaration of non-infringement and separate petitions to revoke two patents, EP (UK) 0 221 570 ("Palmaz I") standing in the name of Dr Julio Palmaz and EP (UK) 0 335 341 ("Palmaz II") in the name of Expandable Grafts Partnership. There was a counterclaim for infringement in the two actions for a declaration. All the actions have come on for trial before me together with an application by motion to amend the specification of Palmaz I. Having regard to the counterclaims for infringement, the onus on all substantive issues was on the defendant or respondent patentees, and by agreement the action followed the customary course of a patent action, being opened by the patentees. I shall refer in this judgment to the plaintiffs, petitioners and alleged infringers as Boston, and to the defendants and respondents as the patentees.

2. The patents in suit are concerned with so-called coronary stents. Coronary stents are used as an adjunct to a remarkable surgical procedure called percutaneous transluminal coronary angioplasty in the treatment of coronary heart disease. I have taken the following largely from Professor Cumberland's words, unchallenged on this introduction to the subject. It is necessary to explain that the aorta is the vessel which carries oxygenated blood from the heart to the systemic circulation. The coronary arteries are two arteries (referred to as the left and right coronary arteries) which lead from the aorta and distribute blood to the muscle of the heart itself. The left artery has two branches.

Coronary heart disease is caused by a narrowing or blockage of the coronary arteries which supply the heart muscle with blood. Arterial walls have three layers: a very thin inner layer called the intima, a middle layer called the media which consists of muscle, both maintaining the structural integrity of the artery and allowing it to contract and dilate, and an outer layer called the adventitia, which is a loose layer of connective tissue. The usual cause of arterial narrowing is atherosclerosis, in which there is a gradual build up of fatty material in the inner layer of the artery wall, followed by deposition of fibrous tissue to produce a plaque which protrudes into the channel (lumen) of the artery. This narrowing is called a stenosis. As the lumen becomes progressively narrowed, the heart muscle fed by the artery concerned becomes deprived of blood when demands are made of it, for example during exercise. The patient may then complain of angina, which is typically a crushing or constricting sensation in the chest and which may spread elsewhere, for example in the left arm or neck. If the protective fibrous cap on the surface over the fat laden core (the so called atheroma) breaks (there is much on going research into the causes of this break or 'fissure') the platelets in the blood stream adhere to the roughened exposed surface and a blood clot forms. The patient's angina may worsen or suddenly appear at this time (unstable angina). If the lumen of the artery suddenly closes off, blood flow ceases and the heart muscle dies, resulting in a heart attack (myocardial infarction).

2. At the priority date of Palmaz I, there were essentially three methods of treating coronary heart disease: drugs, coronary bypass surgery and angioplasty. Drugs may be used to relieve the symptoms of angina by relaxing the muscle of the artery wall, which improves the supply of blood to the heart muscle. They may also be used to make the heart beat less forcefully, so reducing its workload. Clot-dissolving drugs may be used, as may anti-platelet drugs, which reduce the tendency of the platelets (red blood cells) to adhere to the plaques.

3. If drugs alone are insufficient, and there is narrowing and blockage in several arteries, the patient may undergo coronary artery bypass surgery, in which vessels from elsewhere in the patient's body are used to bypass the problem by connecting them to the artery beyond the blockage. I understand that this surgery can relieve angina, and may be successful for many years if the grafts remain open. It will be apparent that surgical intervention requires general anaesthetic, is expensive and time-consuming and involves major trauma for the patient but the evidence was that mortality is low (~2%).

4. Finally, angioplasty. This technique was initially developed in the early 1960's but its application to the coronary arteries became possible after the development of a balloon which, when inflated, was strong enough to dilate an arterial stenosis in a coronary artery. One Andreas Grüntzig in Zurich developed a catheter with a relatively non-elastic sausage shaped balloon near its tip. The catheter had two channels: one for introducing a guide wire along which the balloon catheter could be passed and the other channel to inflate the balloon with fluid at a high pressure (between 6 and 12 bar, that is about 6 - 12 atmospheres. The pressure in a car tyre is about 2 - 2 1/2 atmospheres). The procedure was carried out first in the peripheral arteries, and then extended to the coronary arteries. The first percutaneous angioplasty was performed in September 1977. It is referred to as percutaneous because the catheter is passed through the skin into an artery.

5. The most recent description of spring-type stents before the priority date to which I was referred was the paper by Maass and others (Bundle 3 page 116) in Radiology 1984; 152: 659-663, who describe the insertion of these prostheses in dogs and calves. The springs are wound up, inserted using a special device into the artery, and released. In unwinding, the device expands and so supports the walls of the artery (Figure 3). Although such devices had been proposed, none, so far as the evidence went, had been used in humans, and it is plain that although interesting to investigators there was no question of their being well known, either to clinicians or to the manufacturers of medical equipment of this type by November 1985.

6. Palmaz 1 was applied for in the European Patent Office ("EPO") on 7 November 1986. It was granted on 30 January 1991, and was immediately opposed by Boston and another company called Advanced Surgical Intervention Incorporated. They opposed it on various grounds which I shall have to consider in more detail hereafter, but in the event the opposition was successful, and the patent was revoked by order of the Opposition Division given on 16 June 1993. The patentee appealed against this decision, and the appeal was heard by Technical Board of Appeal 3.2.2. During the course of the appeal, the patentee submitted various so-called requests, that is, alternative forms of claim which it submitted met the grounds of opposition and resulted in a valid patent. The procedure is described in the decision of the Technical Board of Appeal:

7. An application for a European Patent (UK) and the granted patent derive their effect in the United Kingdom from the provisions of sections 78 and 77 respectively of the Patents Act 1977. I am concerned with the provisions of subsections 77(4) and 77(4A):

6. Palmaz I was held to be invalid by the Opposition Division. The order revoking the patent was made on the 16 June 1998.

By Article 68 EPC,

8. Notice of Appeal was given, and in common with many jurisdictions the effect of this appeal was automatically to suspend the decision appealed against (Article 106(1) EPC). Thus the order for revocation did not take effect. The Technical Board of Appeal set aside the order of the Opposition division (paragraph 1 of their order, page 32 of decision T 0818/93). It seems to me that this provision of the order must have deprived the decision of the Opposition Division of all effect. Paragraph 2 of the order is in the following terms:

10. It appears from paragraph 111.02 of Singer, The European Patent Convention (Revised English edition by Lunzer) that this is a standard form of order. As Singer observes, frequently an appellant does not provide the Board with a satisfactorily amended description, and therefore the order has to indicate that the provision of a suitable amended description is a matter which must be dealt with by the first instance. Singer also points out that the decision "normally gives a clear indication of the scope of the claims in the application or patent which in its view ought to be maintained".

8. Finally, it is necessary to observe that the powers of the opposition division, to which the case was remitted, are set out in Article 102 EPC:

11. By Rule 58 of the Implementing Regulations:

9. The final form of the specification was thus left undecided by the Technical Board of Appeals. On 6 May 1998, there was a hearing at the EPO before the Opposition Division, who decided to accept a form of specification slightly different from that which was in issue before me. This decision was immediately appealed by Boston, and this appeal has the effect of suspending the decision appealed against.

10. I have to decide what the form of the patent is now. Does it include the amended claim? Boston say that the remark by Singer which I have quoted above suggests that the amendment only takes effect by way of regrant, when the contents of the patent are approved by the patentee and the requisite fee paid. Until this happens, the old, invalid form of the patent is the one in force. The patentees point out that Art 103 contemplates that the amended specification and the mention of the decision of the Opposition division have to be published together. They submit that the decision to amend the patent is final, even if it comes in instalments. I do not think this is right. I believe that the correct view is that we know that the patent with the amended claim is valid: and we know that it will be granted, provided that the first instance can produce a form of specification which is satisfactory both to the opponents and to the patentees. The best I can say is that I know the form of the claims in the patent when regranted. What I do not know is the form of the specification, although it is likely to be that which has been accepted by the Opposition Division. Until regrant, the order for revocation is contingently alive, and will operate if the patentee withholds his consent to the form of the specification and does not pay the fee for reprinting.

14. So the description counts. It seems to me that it is quite wrong that the form of the description should still be unclear nearly two years after the opposition was dismissed on the basis of the amended claims.

12. My conclusion is therefore that the now current form of Palmaz I, from which the patentees derive their rights, is the unamended form, including the unamended claims. Since it is conceded that unamended claim 1 is invalid as it stands, it must be amended in these proceedings if it is not to be revoked.

13. I should add that the problems with which I have been confronted would, no doubt, be avoided if these proceedings had been stayed until the proceedings in the EPO were complete. No doubt the proceedings would then be legally tidy, but from every other point of view such a course would, in my view, be unacceptable. Rights are conferred on a patentee by grant of his patent. If there is infringement, he should be in a position to take proceedings as soon as he is able to do so, and it is in the public interest that both he and the alleged infringer should know where they stand as quickly as possible. The possibility of two sets of hearings in opposition proceedings with the attendant delays is insupportable. It is to be observed that Palmaz I was applied for in November 1986. Eleven years later, more than half the way through the life of the patent, the ultimate form of the specification is still unclear. Prosecution of the patent lasted 4 years. Opposition has lasted, and is still not complete, from 31 January 1991 to date, a period of 7 years. Courts of the contracting states are left in the unenviable position of having to decide the position of the parties to litigation while proceedings in the EPO slowly advance. Opposition proceedings are in their nature supposed to be speedy (the EPO does not even regard the proceedings at first instance as judicial proceedings), and I believe that the present position does the present system of examination and opposition no credit.

14. The patentees in the present case have applied to amend both claims and specification. That application to amend has been opposed on substantial grounds. The amendments sought substantially duplicate the amendments sought or already made in the EPO. For the reasons which I have given above, I am satisfied that before regrant the specification and claims are in their original forms. Regrant cannot take place until the specification and claims are in their final form and have been accepted by the patentee. Accordingly, the patentees are right to have applied to amend. The real question is the approach which I ought to take to the application to amend.

16. Before considering the exercise of the discretion to refuse amendment, I believe it is helpful to consider the consequence of an amendment for the patentee. For the purposes of this discussion, I take as my starting point the provisions of subsection 62(3) of the 1977 Act, which is a general provision restricting recovery of damages where the specification has been amended:

17. In the 1949 Act, the corresponding provisions were found in Section 62, subsection (2) of which was as follows:

20. As for the effect of refusing amendment of a partially valid patent, Pearson LJ thought it difficult to see what would be the ultimate result of a total refusal to allow any amendment: "I suppose there ought to be brought about in some way a revocation of the patent." So far as the 1977 Act is concerned, the position is perhaps clearer. The principal ground of revocation (paragraph 72(1)(a)) is concerned with the non-patentability of the "invention", which is defined by section 125 to be that which is found in "a claim" of the specification. Subsection 72(4) deals with partial validity expressly and seems to make it clear that either the patent is amended or it is revoked, once invalidity is shown but it will be observed that it avoids all reference to claims. This is because a single claim may specify more than one invention (subsection 125(2)) and may cover matter having more than one priority date

17. It should also be observed that the propriety of an amendment made under the 1977 Act can be investigated (see paragraphs 72(1)(d) and (e) of the 1977 Act) whereas under all the preceding provisions for amendment (going back to the statute of William IV) the amendment, once made, was in the absence of fraud final and conclusive for all purposes, save that recourse could be had to the original specification for the purposes of resolving ambiguities in the amended document (see for example section 31(2) of the 1949 Act).

18. Finally, the specific restrictions on permissible amendments contained in section 76 appear to have been derived directly from Article 123 of the Convention, but they are not expressed to have "so nearly as practicable the same effects in the United Kingdom as the corresponding provisions of the European Patent Convention" by subsection 130(7) of the 1977 Act.

20. It is not clear to me that there is quite such a clear distinction between an amendment by way of deletion on the one hand and a validating amendment on the other, particularly now that both subsection 72(4) and section 63 eschew all reference to claims. The reason for this is no doubt that the Act explicitly contemplates that a claim may comprise more than one invention, and contain matter entitled to different priority dates. I accept, of course, that there is still a class of amendments which introduce into the claims for the first time features which have only been found in the body of the specification, but I think that it is now legitimate to pose the question whether the 1977 Act requires the different types of amendment to be considered differently: and the extent to which it is still necessary to carry out a more or less detailed analysis of the patentee's conduct in considering whether to accede to an application to amend.

21. In the present case, the patentee submits that even if amendment of the patent is a theoretical possibility (because the amendments made during opposition have not yet been finalised and regrant of the patent has not taken place), nonetheless the form of the claims is known; the form of the specification has been approved, albeit appealed, and the EPO will not permit wider discretionary factors relating to the patentee's conduct to influence its decision whether or not to amend. Hence, it is said that if there are factors which would otherwise affect the court's discretion to refuse amendment, they should not be considered here, since the patent will inevitably be amended. Boston's answer to that is that if the patent is not amended, and there are discretionary reasons for refusing amendment, then it ought to be revoked since the existing claims are undoubtedly invalid. Whether or not the EPO grants a patent, it ought not remain in force as a European Patent (UK).

23. Dr Palmaz personally published the embodiment of Figs 1A and 1B of the patent, and he admits it. So the case is at least potentially in the class of cases in which Jacob J contemplated that the EPO might be justified in refusing amendment. But it has done no such thing, although it has invalidated the claim on the basis of what it accepts is Dr Palmaz's own disclosure. It does not seem that the EPO had any regard at all for what we should consider relevant discretionary considerations. Neither counsel drew my attention to any case in which the EPO has ever refused an amendment in the course of an opposition on grounds that an English lawyer would recognise as discretionary, and there are many tens, if not hundreds, of oppositions in which amendments are offered every year. Indeed, as Mr Waugh submitted it is strongly arguable that there is no discretion to refuse amendment in the EPO. If this is so, the anomaly becomes more serious.

24. Approaching the matter as if it were free from authority, I think that it would be wrong for this jurisdiction to impose requirements on patentees seeking to amend which are substantially different in principle from those which are imposed by the EPO on patentees seeking to amend in the course of opposition proceedings. It seems to me that it would be bizarre if amendments effective in all contracting states would be permitted in the course of opposition proceedings by the granting authority but that the same amendments should be refused in the United Kingdom in revocation proceedings, not because of a different appreciation of the legal effect of the relevant provisions but because the court in the United Kingdom considers the patentee's conduct to be in some sense blameworthy. Such an approach would lead to capricious results, depending upon whether the EPO or the UK court disposed of the revocation proceedings and the associated application to amend first.

26. The patent is entitled "Expandable intraluminal graft, and apparatus for implanting an expandable intraluminal graft". It is plainly addressed to a manufacturer of such devices, that is to say, a manufacturer of medical prostheses and the like. Such a manufacturer can be expected to have all the necessary engineering skills to make small articles to a high degree of precision and reliability for implantation in the human body. Such a manufacturer will either have in house, or will be able to call on, clinicians.

27. The specification acknowledges (column 1 lines 30-35) expanding stents of the type which I have illustrated at Figure 3 above, and refers also to "expanding stainless steel stents formed of stainless steel wire in a zig-zag pattern". This is a reference to the so-called Gianturco Z-stents, which are illustrated in Figure 4. This figure shows the stent in the sheath in which it is delivered on the left, and the fully expanded stent on the right.

26. The specification continues, in a criticism which I believe the experts on all sides accepted:

29. There is no suggestion at this stage that the specification is concerned only with prostheses for use in the vascular system. The specification then describes balloon angioplasty as "another alternative to conventional vascular surgery", suggesting that the procedure causes fissuring of the intima, which may cause a flap of material to be formed. It proceeds:

31. Having dealt with other problems of balloon dilation in a passage which it is not necessary to discuss, the specification continues at column 3 line 27 with a recital of the claim.

28. The specification describes the invention with reference to four figures. The grafts of both pairs of figures are said to be usable for many purposes.

33. The invention is said to be quite generally applicable.

29. Figures 1A and 1B show essentially a wire mesh construction which is intended to be placed over a balloon (for example an angioplasty balloon) and introduced to the site of the lesion, and then expanded using the balloon. After expansion using the balloon, the balloon is withdrawn, leaving the prosthesis in place.

30. Thus the steps of balloon angioplasty with an expandable stent can be illustrated like this:

34. The predilation step is not described in the patent. This illustration is not intended to reflect the teaching precisely, but to give an indication of how the technique works.

31. The second embodiment (Fig 2A and Fig 2B) is described at column 8 line 17ff. This is described as follows:

36. There are a number of points to observe about this passage. First, the embodiment of Fig 2A may have non-integral joints, or it may be formed by a technique such as electromechanical or laser etching, which are described as conventional. Second, the graft can be described as a wire mesh tube. Third, no distinction is drawn in terms of inventive concept between this embodiment and that of Fig 1A: the differences are differences in manner of construction. Thus, viewed quite generally the invention consists of a wire mesh tube designed to be fitted over a balloon such as an angioplasty balloon and expanded by the balloon at the site of a lesion. The expansion is permanent and the stent is plastically deformed.

32. The embodiment of Fig 1A and Fig 1B was admittedly published before the priority date by Dr Palmaz. I shall have to consider the extent of this disclosure, and its effect but it is sufficient at this stage to observe that the claim as originally filed is not said to be valid, and the patentees rely upon the claim which has been accepted by the EPO. In order to show the form of the amendment, I set the claim out below, showing the amendments approved by the Technical Board of Appeal which the patentees also seek to make in the present proceedings. The features of the claim are labelled for reference, and this labelling was used at trial to identify the features.

37. An expandable intraluminal vascular graft or prosthesis (70) for a body passageway, comprising:

33. The application to amend can now be considered. I have already expressed the view that the patent has not yet been amended, notwithstanding the order of the Technical Board of Appeal, but that it has not been revoked, because of the same order. The application to amend the claim I have set out above. The amendments to the description which are proposed are very extensive. Their objective is to acknowledge that an embodiment which corresponds to Figs 1A and 1B and which is woven of cylindrical stainless steel wire silver-soldered at the points of crossover is prior art, and to reduce the scope of the claim to cover a prosthesis generally corresponding to some only of the embodiments generally indicated in Figs 2A and 2B. The embodiments which are sought to be covered are those in which the Fig 2A/2B configuration is generally used but in which the generally longitudinal bars are formed to be integral with the generally circumferential bars.

34. I consider that in the exercise of my discretion the determining factors which I should take into account are (1) the claim as proposed to be amended is that already approved by the Technical Board of Appeals and (2) the amendments to the specification are (with immaterial variations) the amendments which have now been allowed by the opposition division. I think that I should only refuse the amendments if they result in a a specification which is invalid, that is, if the invention remains old or obvious, or if the amendments offend against section 76.

36. The re-amended Particulars of Objections take two (not three, as Boston suggest) objections to the amendment of the claims (Particulars (b) and (c) under paragraph 4). Essentially, what is said is that the matter sought to be introduced into the claim and into the specification to distinguish the invention from that originally disclosed is not derivable from the specification as filed. The argument as finally advanced was that the embodiments of Figures 1A/B and 2A/B were not described as patentably distinct in the specification prior to amendment. To the extent that the amendments to the claim succeed in distinguishing these embodiments, matter has been added.

37. I think that this approach to subsection 76 (and Article 123) is not quite correct. If the specification discloses distinct subclasses of the overall inventive concept, then it should be possible to amend down to one or other of those subclasses, whether or not they are presented as inventively distinct in the specification before amendment. The difficulty comes when it is sought to take features which are only disclosed in a particular context and which not disclosed as having any inventive significance and introduce them into the claim deprived of that context. This is a process sometimes called "intermediate generalisation".

38. Features H and I of the amended claim, as I have set them out above, introduce two particular features into the claim. The first is that the bars have a uniform thin cross-sectional configuration. In the specification before amendment, the elongate members could be wires (col 4 line 16) or bars (col 4 line 20) "fixedly secured to one another". It is clear (column 7 lines 40-2 and column 8 lines 28-9) there is no restriction on the shape of the bars: the limitation of feature H does not seem to me to add matter not in substance disclosed in the specification.

39. Feature I is in a different category. The second bars "extending on the circumference of a circle whose plane is perpendicular to longitudinal axis" of the stent are disclosed only in the context of a construction in which the adjacent first bars are parallel to that axis. There is no disclosure of the significance of second bars having such a position in the stent: it is not suggested that this is any part of the inventive concept. The specification does not disclose the use of such bars with first bars which are not parallel to the axis of the stent. It seems to me that this amendment represents the selection of a particular feature, whose significance is nowhere disclosed, and its incorporation into the inventive concept shorn of its original context. This feature is for the first time suggested to have technical significance whether or not in combination with bars parallel to the axis of the stent, and in my view this is an addition of matter to the specification.

40. In paragraph 3 of the reasons for the decision of the Technical Board of Appeal, the Board say that the features of the claim introduced by this amendment are directly and unambiguously derivable from the graft illustrated by figure 2A, so that the subject matter of claim 1 does not extend beyond the content of the application as filed. Because the application does not disclose circumferential second bars otherwise than in the context of longitudinal first bars, I disagree.

41. I consider that the corresponding amendments to the body of the specification are open to the same objection. Thus, if the patent is amended in the manner sought, as it will be, it will be invalid, and I refuse the amendments sought in these proceedings. I will, however, consider the remaining issues as if this amendment had been successfully effected.

42. All references to columns in this section of this judgment are references to the specification as proposed to be amended at tab 2 of the bundle of specifications. The dispute turns on features B, H and I. In fact, there was no real dispute as to the primary meaning of the words of the claim. The scope for any more elaborate approach to the characterising features of the claim (features H and I) is very limited, since the only support for these features is in figures 2A and 2B, and there is no discussion of their significance in the body of the specification. It should however be observed that the claim is not limited to a stent suitable for any particular body passageway, and in particular is not limited to stents suitable for the coronary arteries, although this has no doubt been the main area of application of such devices.

45. In general, the claim describes the article in purely geometrical terms. The wall surface is the outer surface of the device (feature A). This surface is formed by intersecting members, which can be identified as falling into two types (feature B). Boston submit that the first and second members must be distinct and recognisable, and suggest that this is confirmed by columns 11 and 12, and the references to Fig 2. I do not think that the body of the specification helps: the meaning of the feature seems to me to be clear as a matter of language. If the two sorts of member are not distinct, the only feature which might prevent the claim from covering the acknowledged prior art of Figs 1A and 1B is the rectangular section (feature H) and (possibly) that the second bars do not extend on the circumference of a circle (feature I). I conclude that the two types of feature must be recognisably distinct.

46. Features C, D and E do not call for separate comment. Feature F seems to be redundant, in that it is difficult to conceive of a construction in which none of the elongate members (of both types) do not deform when the balloon expands

47. The patentees suggest that in context the two uses of the word"thin" have different significations. The first use ("thin bars") is to exclude constructions in which the width of the bars in the wall surface is not small; the second use is requires the bars also be thin in a radial direction. It is said that it is necessary to attribute these two meanings in order to exclude the acknowledged prior art US-A-3,657,744, referred to in these proceedings as Ersek. (The passage acknowledging Ersek is inserted by amendment at column 4, in accordance with the EPO practice to which I have referred.) This passage says that Ersek, which before expansion consists of a tube in which staggered parallel slits is formed,"also [discloses] a plurality of integrally formed intersecting elongate members having a uniform thin rectangular cross-sectional configuration". The argument is that the requirement for thin bars in feature H is for the purpose of excluding Ersek. Thus, while in Ersek the bars are thin in a radial direction, the use of the words in the claim is to exclude Ersek-type embodiments in which the bars are not also thin in the dimension lying in the surface of the wall.

48. I consider that this is a far too elaborate approach to the question of construction, which is not a process which involves submitting the claim to"the kind of meticulous verbal analysis in which lawyers are too often tempted by their training to indulge." First of all, it is far from clear that the purpose of using the word"thin" twice is to exclude Ersek. Second, it is not clear that the bars in Ersek are not also thin in the surface of the perforated tube.

49. Boston, on the other hand, submit that the words mean that the bar is to be rectangular, not square, and that their width in the surface of the tube is to be greater than their depth in the radial direction, and that both sets of bars are to have the same dimensions. They submit that this is consistent with the representations made by the Uexküll & Stolberg, the patent attorneys acting for the patentees, to the European Patent Office in their letter of 20 July 1995. This is not a legitimate approach to construction either. Accordingly, I prefer the view that the words are used entirely generally to relate the general dimensions of the bar to its required behaviour when it deforms. They do not exclude bars which are generally square (a square is an equilateral rectangle) and they do not exclude constructions in which the bars, while thin, have a radial dimension (depth) greater then their dimension in the surface of the tube (width).

44. Feature I presents no particular problems. There are four requirements: one of the two sets of distinct bars is circumferential; each pair of the other bars must have at least two circumferential bars joining them; each of the circumferential bars joining the pairs of other bars must extend only between the pair of bars which they join; and the circumferential and other bars must be integral. I use the word 'other' rather than 'longitudinal' because there is no requirement that the other bars are generally parallel with the axis of the stent.

45. The alleged infringement is known as the NIR stent. This stent is manufactured in an unexpanded form. It may be crimped over the balloon, and then expanded. As shown in the manufacturing drawings it looks generally like Figure 6 and Figure 8. As crimped on the balloon for use, it looks like Figure 7, and as expanded, like Figure 9.

It appears that this stent has essentially two components. At trial they were referred to as vertical loops and horizontal loops or U-shaped and C-shaped loops. The defendants said that the purpose of the design was to give flexibility, a question to which I shall return when I consider Palmaz II. Mr Richter, the designer, gave his explanation of the way in which they worked under cross-examination. At page 1015 he gave the following answers.

85. Mr Silverleaf QC for the patentees suggested that I should regard Dr Richter's evidence as entirely self-serving. He said that Dr Richter was a man who had a huge vested interest, both technically (as the progenitor of the allegedly infringing design of stent) and financially, since his business is built upon the allegedly infringing stent. I reject this submission. I consider that Dr Richter was not seeking to mislead me. What his evidence reveals is that the alleged infringement works in a rather subtle way.

46. Essentially, the structure of the stent is uniform, being built up of identical unit cells. These cells may be viewed as composed of one horizontal, or C-shaped hoop and one vertical hoop, or U-shaped hoop. On expansion, both sets of hoops expand. The effect of the expansion of the C-shaped (longitudinal) hoop is that the radius of the stent increases. The effect of the expansion of the U-shaped (circumferential) hoops is that the longitudinal contraction of the stent which would otherwise occur is substantially decreased.

47. The arguments advanced in support of the allegation of infringement have varied from time to time. At Bundle X tab 1 is Mr Silverleaf's coloured drawing, which is intended to illustrate his first submission on infringement (Figure 10).

86. Mr Silverleaf submitted that the pink members were the connector members, which he said did not matter for the purposes of infringement, the grey areas represented the second bars and the white members represented the first bars. I fail to see how it could be said that the grey areas intersect with the white areas. Feature B is not present, and it is not possible to satisfy the requirements of feature I.

48. Alternatively, Mr Silverleaf submitted that the NIR stent consisted of slotted tube segments (the C-shaped hoops) joined by U-shaped flexible bights (the pink parts). On expansion, he submitted that the result is a diamond-like lattice (Figure 9) He said that the longitudinal struts in that lattice are formed from the longitudinal members of the slotted tube sections (ie, the white sections) and the pink portions. The circumferential portions are the"ends of the slotted tube portions", that is, the grey portions.

49. Dr Clyne, on the other hand, identified 11 distinct portions of the NIR stent (see section 4 of his first report). Under cross-examination, he seemed to find the second of the connecting (circumferential) members required by the claim in the connection through the white and grey parts all the way round the structure (page 522)

102. Dr Buller, the patentee's principal expert, also accepted that this approach to construction depended on identifying one of the relevant connections as being formed by the rest of the structure (page 956):

51. The bars are, however, thin. However identified, they are not uniform. Their shape is as follows:

110. It is submitted that these are neither thin bars nor a thin rectangular cross-section. If this were the only point on infringement, I would have no hesitation in rejecting it. These bars are generally square. It was suggested that a true (mathematical) rectangle would cause twisting: but twisting is not excluded by the claims. There was some dispute, which was inconclusive, as to whether twisting was beneficial or not; but since the specification does not suggest that the claimed structure aims either to achieve twisting or to avoid it, it seems to me that this consideration is irrelevant. If it had been possible to identify the first and second intersecting elongate members, they would have been thin bars, having a thin rectangular cross-sectional configuration. My view that it is sensible to call the cross-section square is reinforced by the fact that Medinol, the manufacturer of the NIR stent, do so themselves: see the Handbook of Coronary Stents (2nd Edn) page 133.

52. It follows that the NIR stent does not infringe any claim of Palmaz I.

53. I should add that in support of his argument of infringement Mr Silverleaf attempted to deploy an argument that the NIR stent had been developed from the Palmaz-Schatz stent (which consists essentially of two Palmaz stents according to Figures 2A and 2B connected between the castellations at the end of the stent by members which are at an angle to the axis of the stent). He said that a Medinol patent specification (No WO 9603092: bundle X tab 29) revealed that the NIR configuration was mechanically equivalent to a construction with square castellations in the horizontal loops. He cross-examined Dr Richter, the designer of the NIR stent, about it. I should make it clear that I do not accept that the NIR stent is, in any reasonable sense of the word copy, a copy of the Palmaz-Schatz stent. Whether or not it was designed with an eye to that stent is quite irrelevant to the question of infringement.

54. The validity of this patent is challenged on the basis of the disclosure of Ersek, to which I have already referred, and on the basis of two disclosures by Dr Palmaz himself. It is admitted that Dr Palmaz did disclose the stent of Figs 1A and 1B before the priority date; Boston say he disclosed the stent of Figs 2A and 2B as well, or that the construction of Figs 2A and 2B was obvious in the light of his disclosures. The two disclosures now relied on are (a) non-confidential discussions between Dr Palmaz and a company called Shiley (the"Shiley disclosure") and (b) a public presentation by Dr Palmaz at the November 1984 conference of the Radiological Society of North America (the"RSNA disclosure"). Boston also rely upon obviousness in the light of a subsequent publication by Dr Palmaz entitled"Palmaz Expandable Intraluminal Graft: Preliminary Study Paper" but Mr Kitchin accepts that this adds nothing to the admitted disclosure of the Figs 1A and 1B embodiment.

55. Mr Silverleaf conceded before me that the discussions with Shiley were not in confidence. This concession was fully justified. Confidentiality in the negotiations with Shiley had been expressly excluded by the Disclosure Agreement package which was sent by Shiley to Dr Palmaz and returned by him, duly signed in November 1992. (see Bundle 13 tab 13 page 76).

56. Both in his witness statement and under cross-examination Dr Palmaz described in some detail the making of his invention. He started by making a stent out of stainless steel wire, which he painstakingly silver-soldered at every joint. By 1980, he had written a document, which was referred to before me as the 1980 monograph. This document was exhibited by Dr Palmaz to an affidavit sworn in the United States Patent and Trademark Office in the course of a so-called re-examination of the corresponding US patent. In this affidavit, Dr Palmaz was concerned to show priority of invention over the filing of another patent by another inventor, which had been applied for before Dr Palmaz had applied for his patent in the US. In the affidavit, he says that in July 1980 he had made his prototype graft, and had expanded it with an angioplasty balloon catheter. He had then sought funding from a number of companies, including Shiley. In order to interest these companies in his invention, he had sent them copies of the 1980 monograph.

57. The 1980 monograph starts with a historical review of balloon angioplasty and an assessment of the results reported for this procedure. It proposes (page 5) an alternative to balloon dilatation in the following terms.

113. The monograph proceeds to give directions for placing the graft over an angioplasty balloon, and introducing it in the conventional manner through the femoral artery. It concludes with suggestions as to the development of the experimental project, including experimentation of different wire structures by changing mesh density, wire diameter and wire material to establish adequate dilating pressures, resistance to expansion and memory of the mesh.

58. Dr Palmaz accepted under cross-examination that by the time he wrote the 1980 monograph he had conceived of a slotted tube device, and had shown it in a picture: see pages 157 and 158:

121. Dr Palmaz's cross-examination then covers the visits which he made to two manufacturers. The first was called Vascor. He went to Vascor with a presentation which included a slide. This slide looked like the figure which I have set out above but in white on a blue background. He showed the slide to Vascor, and he gives this important answer:

128. The disclosure to Shiley is covered in his cross-examination on pages 175 to 180. At the time of his discussion with Shiley, Dr Palmaz said that the slotted tube was in his mind as a way of making the product (page 176 line 4). He knew that a soldered wire construction was difficult to make (page 176 lines 11-12). He says that he cannot now recall whether he talked about the slotted tube to Shiley at all (page 176 line 21), and at page 179 he says that he does not now think that he got to the slotted tube, because he did not know how to make it. I should at this point say that Dr Palmaz seemed to me to be a very fair witness. He did his best to deal with what must have been extremely unwelcome suggestions and I was struck by the fact that he never flatly denied that he disclosed the idea of the slotted tube to Vascor or to Shiley. His answer at page 179 line 25 was characteristic. I think that he thought that he might have disclosed the idea to Shiley, but was not satisfied that he had.

59. I have to consider the probabilities. These meetings were important in the pursuit of funds. The silver-soldered wire mesh embodiment was admittedly difficult to make, and was unattractive because it contained many materials and was far from uniform, and potentially thrombogenic for this reason. The slotted tube was uniform. I think that it is much more likely than not that Dr Palmaz gave these attempts to find funding his best shot. The meetings included a presentation, and it is normal and natural at such a presentation to explain everything on the paper or on the slide. I find that Dr Palmaz showed and described a slotted tube embodiment of his idea to Vascor and to Shiley. Once the article shown in the picture is described as a slotted tube, it was not suggested that there was any difficulty in making it (if there were, the specification, which is silent on the question save to say that the slots are formed by a conventional etching process, such as electromechanical or laser etching, would be insufficient) and I find that the slotted tube embodiment of the 1980 monograph was disclosed as such to Shiley. This disclosure was an enabling disclosure, and claim 1 as proposed to be amended is anticipated by this disclosure.

60. Mr Silverleaf emphasised, correctly, that if the right hand side part of the picture was not described as a slotted tube, I should be very cautious before coming to the conclusion that a slotted tube was disclosed. Whilst I have found that Dr Palmaz did so describe this drawing, I must deal also with this submission, since it cannot be disputed that the drawing, and probably the slide also, were shown to Shiley, even if they were not explained.

129. In considering this question, hindsight represents a real danger. The evidence relating to what was disclosed by the right hand side of the drawing in its context absent a statement that it showed a slotted tube was not entirely satisfactory. Professor Cumberland, an excellent witness, gave evidence in his report that the figure disclosed a slotted tube. He was cross-examined at length on this (pages 1147 - 1154). He accepted that he had seen the Palmaz stent (indeed, had used it) before he saw the figure. Dr Buller gave evidence that one could make a wire structure which looked generally like the unexpanded part of the right hand figure. The text is entirely silent as to any method of making the article which does not involve welding or soldering. There is no suggestion of an integral device.

61. On the evidence, I do not believe that if the drawings only, without explanation other than that contained in the 1980 monograph, were shown to Shiley, the idea of a slotted tube was clearly and unambiguously disclosed. But it seems to me to be an absolutely obvious interpretation of the figure. The unexpanded object shown in the picture is clearly tubular. It expands to give the mesh on the right - this follows from the explanation of the woven wire device. Thus, it is a tubular stent, which expands to form a mesh and has slot-like openings in it. I am satisfied that a slotted tube is a perfectly natural and obvious interpretation of this figure, and that claim 1 is obvious over this disclosure. I should add that Boston sought to rely upon certain evidence given on behalf of Johnson & Johnson in proceedings in Delaware to show that a machinist accustomed to articles of this kind, who would be part of the team of persons to be considered as the addressees of this specification, would be well aware of the methods of machining needed to make an article of this description (Bendel transcript, at Bundle 13 tab 18). I have already indicated that if such methods of machining are not common general knowledge of the addressee the patent would be insufficient, and I am satisfied that to such a person the interpretation of the photograph as an integral slotted tube would be a very natural and obvious interpretation.

62. By the time that this meeting took place, Dr Palmaz had edited his monograph to produce a new version, which was referred to as the 1983 version. While the figure remains the same, the accompanying text is as follows:

130. The slide which accompanied the RSNA presentation was substantially the same as Dr Palmaz's figure 1 which I have shown above, again with the difference that the background was dark blue and the figures appeared in white.

63. Accordingly, by the time of the RSNA presentation, Dr Palmaz had himself written a description of both embodiments illustrated in the figure. Boston rely upon a short passage in a deposition of Dr Palmaz taken in proceedings in the US, which was put to him in cross-examination at page 196 line 11 to 197 line 26. I think that it is established that Dr Palmaz made a short presentation at the RSNA meeting in question and made use of the slide during the presentation. The purpose of the presentation was to show colleagues what his ideas were and how far they had come. The presentation was one of many at the meeting in question, and was short (Dr Palmaz said about five minutes, and the summary, to be found at X1 tab 1, suggests that with questions the whole presentation took 8 minutes). The summary begins as follows:

131. Obviously there is no mention of a slotted tube. The question is again whether Dr Palmaz explained the slide which he certainly presented, and whether that explanation included a mention of a slotted tube and, if it did not, whether a slotted tube was either clearly and unambiguously disclosed by the slide itself or an obvious interpretation of the slide.

64. In my judgment, it is much more likely than not that Dr Palmaz explained his slide by explaining what both the embodiments which it depicts were. Accordingly I find that he again disclosed the slotted tube as such at the RSNA meeting. If I am wrong about this, I am satisfied, for the reasons which I have already given, that a slotted tube is an obvious interpretation of the figure.

65. The final question is whether the disclosure of the slotted tube is sufficient to anticipate claim 1 as proposed to be amended. On this question, it is necessary to consider the effect of forming a slot in a tube of the type with which I am concerned. The evidence was that slots must be of a finite width: Thus, in the real world a slotted tube will always contain short circumferential bars. The picture of the unexpanded stent plainly shows slots of finite width, and circumferential bars. It is quite plain therefore, that features B, H and I are all present. Feature B appears from the drawing, and the fact that the object is a slotted tube. H follows from the fact that simple slots made in a plane tube will leave members rectangular in section. Feature I appears from the drawings, and the fact that this is a tube with slots necessarily means that the second bars are formed integral with the pairs of first bars.

66. It follows that the claim as approved by the Technical Board of Appeals is anticipated by Dr Palmaz's own presentations and descriptions. I regard it as regrettable that the RSNA disclosure, which took place in the year immediately prior to the filing of the US application from which the patent in suit claims priority, should have this effect, when it would not be available as a prior publication in the United States itself. This problem, which is a familiar one with inventions published in the"grace period" permitted by US law, will continue until there is international agreement as to whether grace periods are desirable at all.

68. Both the experts were agreed that the blobs of solder are clinically undesirable: Professor Cumberland paragraphs 64 to 68 and Dr Buller at paragraphs 71-74. Dr Buller did not give evidence that this would be clear merely from a consideration of the woven article alone. His evidence at page 824 line 15 to 828 line 13 is criticised by Mr Kitchin, who submits that it is not consistent (Mr Kitchin compares particularly the answers at 825 lines 1-14 with the answer at 826 lines 12-828 line 3, and I believe also the answers at 798 line 23 to 799 line 25). I think that this is not a so much an inconsistency as a reluctance to accept that the defects of turbulence and of a risk of electrolytic action caused by the presence of different metals would be apparent to the skilled man and would point him firmly in the direction of a smooth tube (no turbulence) with integral joints (no electrolytic action).

69. Professor Cumberland accepted that the problem of different materials would be overcome merely by welding: 1161 lines 7 - 9. Mr Silverleaf submits that in truth at the date of the original disclosures the stents would have been seen as potentially thrombogenic, dangerous devices, and that the first reaction of the addressee would be to dismiss the whole idea. At any rate, he said, it is clear that there was no way forward which would not involve investigation, at least requiring a research programme: see Professor Cumberland's answer at 1162 lines 14-16. I think that the re-examination of Professor Cumberland shows that he was really considering the investigation of the limitations of the device: see 1202 to 1204, but it also reveals that he would move forward in small steps.

71. Identification of the inventive concept of the patent in suit. This is what is claimed in claim 1, correctly construed. I think that an attempt to gloss the actual words of the claim can lead to error. Thus, the inventive concept includes features H and I, and includes the need for identifiable circumferential bars.

72. Assumption of the mantle of the addressee. I have already identified the addressee as a maker of surgical devices and particularly prostheses suitable for implantation in the human body. Such a person has the manufacturing skills to which I have referred above.

73. Differences between the matter cited and the invention. These may be summarised as a slotted tube rather than a woven wire with welds or soldered joints and the circumferential members in the device.

74. Obviousness of the steps constituted by the differences in going from the matter cited to the allegedly inventive concept. In the end, I have come to the conclusion that if I am wrong about the disclosures made by Dr Palmaz, and if I am wrong about the interpretation that the addressee would place on the right hand side of the figure absent those disclosures, it would be obvious to make a device from a slotted tube. Such a device would not inevitably fall within the claim as I have construed it, since it would not necessarily possess feature I, that is, it would not necessarily have circumferential second bars. If, however, the claim is to be construed so that such bars can be found merely as part of an undifferentiated region where the bars meet, then the claim is obvious, since all slotted tubes on expansion would have such regions. I am not satisfied that a device having feature I as I consider that it should be construed would be obvious.

75. Dr Palmaz's disclosures include balloon expansion of the stent. Thus claim 3 is not valid independently of claim 1 in relation to any of those disclosures.

76. Ersek is entitled"Method for fixing prosthetic implants in a living body". It discloses a device, clearly illustrated in Figures 2 and 3, which are alternative embodiments, of a so-called fixation sleeve, which is intended to act as a means for joining a graft or prosthesis to an existing vessel. Figure 1 shows the insertion of a synthetic bifurcated aortic graft. The two arteries at the bottom of Figure 1 are the iliac arteries. The joins are made by inserting the expandable metal sleeve over an expansion tool, positioning the sleeve at the correct place and then actuating the tool so as to expand the sleeve until it engages with the graft and with the vessel. The sleeve may be made of flat stainless steel in which slits are made (column 2 lines 56 - 69) which is then expanded. The specification recognises that the process of expansion will twist the ribbon-like portions of the expanded material (col 2 lines 69 - 71) but it is said that it is preferable not to flatten the material before it is shaped into a sleeve (col 3 line 1) since if there is no flattening there will be a plurality of narrow projecting edges which embed themselves in the vessel wall on expansion of the sleeve. At column 3 line 19, the specification says this:

78. Ersek has all the features A-H of the claim: this is correctly conceded in the amended specification at column 4. Some evidence was called which was intended to suggest that this concession was wrongly made. This is an unattractive approach to take to a passage which it is sought to insert in the specification by amendment, but is in any event incorrect. Mr Silverleaf submitted in opening that Ersek did not describe a stent. This is of no relevance, since the claim describes the device by reference only to its physical features. The only question is therefore whether it has features H and I. Feature H is present: the elongate members are plainly present, and if they are formed from slotted sheet they must be thin (in the sense in which I have construed this word above) and rectangular.

79. Feature I is a different matter. The sleeve in Ersek is described merely as being made from sheet in which slots have been formed. Any circumferential member can only be as long as the slot which it forms in combination with the pair of elongate members is wide. Certainly no circumferential member is identifiable in the drawings. Mr Kitchin submits that the suggestion that Ersek does not have the circumferential member is untenable, because there is no restriction in the patent on the length of that member. That is undoubtedly true, but there must still be an identifiable member. He points out that that slots or slits cannot be narrower than the minimum that the relevant process permits: he suggests 200u (one-fifth of a millimetre) as a lower limit for electrodischarge machining (EDM) and 100u for photo-chemical etching, on the basis of Dr Clyne's evidence at 445-8 and Dr Richter's evidence at 992. Dr Clyne's evidence was plainly that 200u was an approximate figure. Dr Richter's evidence at 992 was that the slot is as wide as the thickness of the material which is being etched. All this is inconclusive as to the presence or otherwise of identifiable circumferential members in Ersek, and I conclude that if feature I is construed as I consider it should be construed, there is no anticipation.

80. The case is altered if the claim is to be construed so as to cover the alleged infringement, so that the second members can, in effect, be located within the joint regions of the mesh. If this is a correct construction, then Ersek would infringe if made after the date, and accordingly anticipates.

81. The only difference of substance between the claimed invention and the disclosure of Ersek lies in feature I. Thus, if it would be obvious to make the slots or slits in Ersek's starting material of sufficient width that it would be possible to identify circumferential members, the claim is bad. As I have indicated, I find no clear disclosure in Ersek of such members, but it seems to me that to vary the width of the slots is pre-eminently the making of a mere workshop variation, and slots of any substantial width will result in the presence of circumferential members. Claim 1 is obvious in the light of Ersek, but claim 3 is not. Ersek's delivery and expansion apparatus does not render balloon expansion obvious.

82. Palmaz II claims priority from 28 March 1988. The application for Palmaz I is a prior publication. The patent is concerned with the problem of inflexibility of the graft. The notional addressee of the specification is the same manufacturer, having the same ability to call on clinicians and specialists in manufacturing techniques, as for Palmaz I. After setting out the advantages of the balloon-expandable stent in columns 1 to 3, at column 3 line 17 the specification proceeds as follows:

136. At column 4 the claims are summarised, and it is clear from the description of the method of implanting the graft at column 4 line 42ff that the draftsman of the specification uses the terms"prosthesis" and"tubular member" more or less interchangeably. It is clear from a long passage in the specification that the draftsman had in mind that the"tubular members" should be the stents described in Palmaz I.

137. The claim, divided up to identify the different features, is as follows.

83. The claim is made unnecessarily confusing by the reference to"at least one tubular member" in feature B, and the characterising feature F which calls for a plurality of such members, but the meaning, at least in this respect, seems to me to be clear enough. I have already illustrated the alleged infringement, the NIR stent, above. The issues on infringement turn on features C and F.

84. Boston's contention is that the claim requires that the unexpanded graft have two identifiably different components: tubular members and connector members. The connector members are required by feature F to be flexible, and to join the tubular members which do all the work of the stent as a stent, that is, expand and deform and expand the lumen of the passageway. Accordingly it must be possible to identify distinct, slotted tubular members in the NIR stent, which are capable of expanding and performing all the functions of a stent. They base themselves on the long passage in the specification to which I have referred, and the interchangeable use of the terms prosthesis and tubular member.

85. This is a method of construction which must be approached with some care. It is not legitimate to use"stray phrases in the body of the specification" either to cut down or to widen the meaning of the claim since to do so may be either unfair to the patentee (who may be justified in seeking to support a wide monopoly with a comparatively limited disclosure) or to the defendant, who should always be entitled to a reasonable degree of certainty, as the protocol on the interpretation of Article 69 of the European Patent Convention makes clear. Nonetheless, it seems to me that Boston's contention is essentially correct. The claim itself draws what seems to me to be in context a clear distinction between the tubular members, which are too inflexible for use in places in the vascular system where grafts may be required (although the claim does not require them to be substantially rigid) and the flexible connecting portions whose function is to articulate the graft.

144. The requirement for slots in the tubular members seems to be to be straightforward, although a great deal of time was spent on it at the trial. Both plaintiffs and defendants drew the distinction between open-ended and closed slots. This does not seem to me to be relevant. The only question is whether the part of the alleged infringement which is alleged to be represent the thin-walled tubular member can be said to be a member which has a number of slots formed in it. These are ordinary words which are not capable of more exegesis.

86. The patentees said that the parts of the NIR stent which consisted of what I have called the C-shaped members represent the"thin walled tubular member having" a plurality of slots formed therein" and the croquet hoops or vertical hoops as representing the connector members. They reinforced this submission by reference to photographs which, they submit, show that when the unexpanded NIR stent bends most of the bending is accommodated by the vertical loops, the C-shaped members remaining substantially undistorted, and they produced a number of illustrations of which Figure 11 (X26) is an example, which are said to illustrate this. It will be observed that when the unexpanded stent is bent, it is clear (as one would expect) that the decrease in length on the inside of the curve is accommodated by a change in the leg spacing of the vertical (U-shaped) hoops. At the same time, it seems that the C-shaped horizontal hoops are themselves distorted, essentially by twisting further out of the cylindrical surface of the wall of the stent. Dr Richter said this in the evidence which I have quoted above, and I accept his evidence. I should add that the patentees also sought to adduce evidence from Dr Clyne relating to a finite element analysis of the stent, carried out using a particular computer program. I formed the view that the evidence was experimental evidence which was not being adduced using the proper procedural mechanisms, which are intended, so far as possible, to make it clear what fact the evidence is intended to prove and to protect the other party from surprise. I did not prevent the patentees from seeking to put in a late Notice of Experiments, but no application to do so was forthcoming, and so the evidence was excluded.

87. The patentees sought to identify the cylindrical elements constituted by the C-shaped hoops as the tubular members. In Figure 11 above, they say that the boundaries between the tubular members and the connector members are shown by the red lines. During the course of the proceedings, they marked up a drawing (Bundle 5 tab 18 page 141) with an indication of what they considered to be the connecting member: Figure 12.

88. Boston admit that the NIR stent as a whole is a tubular member in the sense of the claim, but deny that the individual ring structures satisfy the requirements of features B and C. Because the stent as a whole is admitted to be a tubular member, it must have first and second ends, and these first and second ends are provided by the wavy ring structures. Feature B of the claim is satisfied. Equally, it is conceded that it has"wall surface". In my view, it is straining language to describe what is essentially a wavy ring as a tubular member having a wall surface of substantially uniform thickness, but whether that is right or not, I am firmly of the view that the ring of C-shaped members is not geometrically an object satisfying feature C of the claim. What is not present is a member in which slots are formed and which is connected to the adjacent tubular member by a connector member.

91. Does the variant have a material effect on the way the invention works?

The principal evidence on this aspect of the case was given by Professor Edelman (a rather combative witness), Dr Richter, the designer of the NIR stent, and Dr Clyne. I think that the evidence as a whole established that the NIR stent was not equivalent in Mr Silverleaf's phrase to a series of napkin rings connected by flexible vertical hoops. Professor Edelman was cross-examined on the manner in which the NIR stent accommodated bending at page 577, and Dr Richter at pages 1003-5. At 1006-7 Dr Richter marked the parts of the C which he said flexed as part of the bending process on Figure 11 where they appear as two circles in the centre of the stent. There was never any effective challenge to this evidence. Dr Clyne accepted in cross-examination that there was deformation of the C-shape members to permit the U-shaped members to bend (pages 534-5). Two further photographs, published in the so-called FINESS study on stent equivalence were also used to demonstrate this point (Bundle 5 page 218).

92. It was common ground that the deformation of the C-shaped members did not, of itself, cause curvature of the part of the ring of C-shaped members, but it facilitated the flexing of the stent as a whole. As I think one would expect, it is misleading to regard the NIR stent as composed of discrete elements: all the parts of the stent co-operate to permit flexing.

93. It seems to me that taken as a whole this evidence demonstrated that the NIR stent construction did not flex in the same way as a stent of the type called for by the claim. I believe that it is clear that this difference is material. Since the claim requires relatively inflexible tubular members joined by relatively flexible connector members, one would expect it is legitimate to look at a commercial embodiment which it is agreed is manufactured according to the claim to see whether it performs in a manner materially different from the NIR stent, and that this difference in performance is due to the variant in construction which I have described. It was clear from Mr Woodall's evidence that lack of flexibility was always a problem with the Palmaz stent, that is, the stent according to figures 2A and B of Palmaz I. The first step towards flexibility was two such stents joined together by a single thin connector bar (the Palmaz-Schatz stent sometimes referred to during the trial as the box-car construction): see Figure 13. This stent falls within the claim. The next stage was a stent like that shown in figures 7 and 9 of Palmaz II, in which there are a number of spiral connecting members. This construction was comparatively rigid. The Johnson & Johnson view of the NIR stent is clearly and strikingly pointed out in the memoranda at X13 and Bundle 7 page 302. These memoranda, the accuracy of which was confirmed by Mr Woodall in cross-examination, point to a design much more flexible, and much more satisfactory than the Palmaz-Schatz stent and the spiral Palmaz-Schatz stent. The improvement was essentially in flexibility. Indeed, it now appears that the Palmaz and Palmaz-Schatz stents have had their day, and are no longer manufactured.

95. Mr Silverleaf also advanced an argument of infringement again based on design history of the NIR stent, as revealed by Medinol patent applications in the name of Dr Richter and others. This is not the correct approach to the question"infringing or not?" which calls for an entirely objective investigation. The whole cross-examination is at page 1027 line 8 to 1032 line 6. Mr Silverleaf proposes a chronology of development which was rejected in terms by Dr Richter, whose evidence Mr Silverleaf submits is shown to be baseless by the timing of the patent applications. I cannot reject Dr Richter's evidence on this basis. The timing of patent applications may or may not reflect the order in which the developments which they record was made and there was no other evidence.

96. The validity of Palmaz II was challenged on the ground of obviousness only. The citations relied on were the original Palmaz stent; certain articulated versions of the Gianturco stents which I have illustrated above (Figure 4) and coiled stents of the type I have illustrated above (Figure 3).

97. Boston submit that by the priority date the common general knowledge which is to be attributed to the addressee of the specification included the procedure of balloon angioplasty; the problems associated with balloon angioplasty (such as restenosis);"stents as a way forward"; and the Palmaz I balloon expandable stent.

98. The common general knowledge has been defined in different ways. It is what constitutes the mental equipment which can properly be attributed to every addressee of the patent. I have no doubt that the procedure of balloon angioplasty was well known to practitioners and the manufacturers of surgical equipment by 1988. The case was certainly opened to me on the basis that practitioners had started to use the procedure in 1980, and was well known by 1988, as were its problems, particularly restenosis and acute occlusion at the angioplasty site during or immediately after the procedure.

99. A considerable amount of effort was expended in an attempt to show that the Palmaz I stent was common general knowledge by 1988. It is noteworthy that by the priority date of Palmaz II there seem to have been no published studies of the use of the stent in human beings. Boston say that stents were regarded as a good basis for further action, and they refer to many publications of experimental work on stents. Most importantly, perhaps, Professor Cumberland gave clear evidence (paragraph 32 of his report) that during 1986 the Palmaz stent became generally known to all those interested in interventional cardiology and radiology. He says that the Palmaz I stent and the Gianturco Roubin I stent were of considerable interest to workers in the field, and that such workers were already well aware of the coiled devices. That may well be so, among those who knew of stents and understood them. It is clear that some were uninterested in them and that others were actively promoting other solutions to the problems of coronary heart disease. Professor Cumberland had written a short review article in 1986 for the Lancet (X33) about the use of balloons, and while he refers to some"futuristic ideas" but stents are not among them (see his cross-examination at 1092-1098 which deals also with other publications at X32 and X34). I would expect that if not only stents, but a particular type of stent, had achieved the status of something of which everybody was aware, it would rate a mention in articles such as these.

100. I suspect that this cross-examination was intended to show that stents were actually seen as an unpromising way forward. This was not established. I believe it is clear that they were viewed a promising new technology. It may be that the fact of the existence of stents was generally known. That does not make stents themselves, their structures and their problems, part of the common general knowledge. No cardiologist could have hands-on experience of them unless he was engaged on clinical trials in humans. I am quite satisfied that the devices were still experimental. I cannot accept that Palmaz I stents were common general knowledge.

101. The relevant pleaded publication is Dr Palmaz's publication in his"Preliminary Study" paper: Palmaz et al., Expandable Intraluminal Graft: a Preliminary Study Radiology 1985 156:73-77. On page 76 of this article appears the following passage:

147. To implant stents in tandem they are either loaded on a single balloon and expanded simultaneously, or they are successively introduced, the catheter being withdrawn and reinserted. The latter alternative is very unattractive if it can be avoided, since it means passing the balloon twice. There was considerable evidence that stents can be detached from the balloon in the course of the procedure. Dr Palmaz accepted it under cross-examination (page 237 to 239) and Professor Cumberland in chief at 1068-70.

148. I was left with the strong feeling that this issue was made to appear more complicated than it really was. Such a feeling is a snare when considering obviousness, and I have gone over the evidence again to satisfy myself that it was not unjustified. It is desired to introduce two short stents, rather than one long one, because the long stent is inflexible. Thus, the two short stents will lie in the artery at an angle to each other, so as more nearly to accommodate the curve. At the same time, the two stents should not be inserted sequentially, if that can be avoided, and they should be prevented from detaching themselves from the balloon. They are difficult to position being very small (one-half an inch or less long) and difficult to see by X-ray fluoroscopy, and inserting two one after the other also doubles the difficulty of positioning them, which is already great. (In this context, I should say how greatly impressed I was by the recordings of some procedures carried out by Dr Buller which I was shown). It seems to me that a single flexible tie of some description between the two stents, so as to help them to be correctly relatively positioned, is absolutely obvious.

102. Gianturco stents

150. The relevant publications are Charnsangavej et al., Stenosis of the Vena Cava: Preliminary Assessment of Treatment with Expandable Metallic Stents, Radiology 1986 161:295-298 (Bundle 3B tab 4) Rosch et al., Modified Gianturco Expandable Wire Stents in Experimental and Clinical Use, a conference paper delivered at CIRSE, Porto Cervo, Sardinia, in May 1987 (Bundle 3B tab 5) and Wallace et al., Tracheobronchial Tree: Expandable Metallic Stents used in Experimental and Clinical Applications Radiology 1986; 158:309-312 (Bundle 3B tab 6). These do not in themselves render the invention obvious. The Wallace paper, which describes two of these expandable stents joined together by a short connecting member (see figures 2 and 3) confirms me in my view that if it is necessary to have two stents maintained at a particular distance from each other yet able to flex, the way to do it is to join the stents together with a flexible strut. It does not matter whether what Wallace shows is a flexible strut or not (see the cross-examination of Professor Cumberland at page 1171): if it is desired to make the connection stiff the strut is stiff, and if the connection is to be flexible, then the strut is flexible.

103. The Rösch article describes two Gianturco stents joined together by joining them with monofilament line. It says the"two-body stent, connected with a monofilament line, is less rigid and presumably allows for some motion, more than a single body, 4 cm long stent." This again illustrates the obviousness of flexibly connecting two shorter comparatively rigid bodies in substitution for a longer, comparatively rigid body, when increased flexibility is required.

104. Claims 2 and 3 of Palmaz II are said to be independently valid. Claim 2 requires the connecting strut(s) to be out of parallel with respect to the longitudinal axis, and claim 3 requires the struts to be coplanar (an inappropriate word) with the tubular member. I believe that this means that they should lie in the same cylindrical surface as the wall of the tubular member. Increasing the non-parallelism of the connecting struts increases the flexibility of the connection over the same number of struts parallel to the axis. This is obvious. Professor Cumberland gave evidence that he considered such an expedient to be obvious from first principles: see paragraphs 87-90 of his report. I do not believe that these passages were challenged directly in cross-examination. In any event, they seem to me merely to reflect common sense. I cannot conceive that the connector strut(s) would be positioned out of the common cylindrical surface. Accordingly I consider that claims 2 and 3 are invalid for obviousness also.

105. Insufficiency was pleaded, I was told, as a squeeze argument on construction or infringement. I am surprised at this, since particular (a) seems to be an old-fashioned plea of inutility. Particular (b) is not a squeeze argument, and fails because it presupposes that it is right to take a simple-minded and literal approach to construction, rather than a purposive construction. The same goes for particular (c) and particular (d) cannot be a particular of insufficiency, since if particular (d) were a good plea any patent infringed by an infringement inventive over the disclosure of the patent would be invalid.