Corevalve Inc v Edwards Lifesciences AG & Anor

Irvine, California is home both to Edwards Lifesciences – the largest artificial heart valve company in the world – and its competitor, CoreValve Inc. I shall call them “Edwards” and “CoreValve”.

Edwards owns European Patent No 0592410. CoreValve says this patent is invalid for anticipation, obviousness and insufficiency, and applies for its revocation. Edwards says that, on the contrary, the patent is valid and CoreValve has been infringing it by supplying its ReValving system; Edwards counterclaims accordingly. CoreValve denies that its product is covered by the claims of this patent, even if it is valid. As an extra line of defence, CoreValve says that it has been supplying the product for experimental purposes relating to the subject-matter of the invention (section 60(5)(b) of the Patents Act 1977).

As I assimilated the evidence and arguments in this case I gradually and increasingly formed the impression that the patent is valid, but that CoreValve’s product does not infringe it. And that is what I hold in this judgment. I must confess, however, that after the conclusion of the argument I had second thoughts: I was side-tracked by a consideration which caused me much trouble, which took up a lot of time in the preparation of this judgment, but which I now believe to be without substance. I shall return to that later. I greatly regret the resulting delay.

The priority date claimed for the patent, and not disputed in this case, is 18 May 1990. Therefore I shall need to consider the state of the art at that date, including what was common general knowledge.

The patent is about artificial valves for implantation in the human body. Such a valve might be used to replace a natural valve. For certain medical conditions it might even be implanted where no natural valve exists e.g. in the oesophagus. But the main focus of the patent is about valves for replacing defective heart valves.

A little introductory terminology may be helpful. Before the date of the patent it was known to assist the patency, or openness, of a blood vessel e.g. a coronary artery by implanting a supporting scaffold called a stent, which may be thought of as a slender wire cage. The stent could be delivered remotely from a blood vessel in the patient’s leg by a technique called catheterisation. In essence, a tube was passed up into the coronary artery and the stent was delivered over an internal guide wire. A small balloon was inflated inside the stent to cause it to expand and the catheter apparatus was withdrawn, leaving the stent in place.

The general idea of the patent is to implant a replacement valve remotely by catheterisation e.g. through a vein, without major interventional surgery, and to leave it in place, so that the patient may lead a normal life. This is to be achieved by mounting the valve on a stent; both the valve and the stent are elastic and can collapse i.e. they can squash inwards in order to be narrow enough to navigate through small passages e.g. arteries. On arriving at the correct site the valve and stent can be re-expanded and left in place.

The heart is a remarkable organ. It took centuries for the world’s best anatomists and physiologists to figure out its mode of operation and I understand that they have not finished even now.

The heart is a pumping device. All vertebrate hearts work on the principle that a muscular chamber, called a ventricle, contracts to drive out blood (during systole) and expands to admit it (during diastole). A pump engineer would say: why not have an antechamber, where returning blood can be collected so that it can be discharged into the ventricle at the right time during the pumping cycle? But nearly all vertebrate animals have evolved one of those, and it is called an atrium.

We mammals (and also birds) have also evolved an ingenious arrangement to overcome the problem that a substantial pressure drop occurs as the blood passes through the lungs. The solution is to have parallel circulation systems or, in other words, a left heart and a right heart. The left heart drives the main or systemic circulation of the body. The right heart drives the pulmonary circulation. This arrangement was evolved in rudimentary form by our amphibian ancestors but in our heart the septum provides a complete division between left and right ventricle.

Consider blood returning to the heart from the main tissues of the body through the veins. It enters the right atrium through the vena cava. From there it passes to the right ventricle through the tricuspid valve. Our pump engineer would rightly say that since the heart is a positive displacement pump it should have valves to prevent the blood flowing the wrong way during the compression stroke. That is indeed what the healthy valves of the heart do; and we can start with the function of the tricuspid valve. During systole the rising blood pressure in the right ventricle causes this valve to slam shut, and so it prevents the blood running back into the right atrium. Instead the rising blood pressure causes another valve to open – the pulmonary valve – so that the blood is driven through the pulmonary artery and thence to the lungs. The function of the pulmonary valve is to stop back pressure causing blood to flow backwards into the right ventricle during diastole.

Having dropped waste gases in the lungs and picked up oxygen, the blood now proceeds – to the left heart, this time – through the pulmonary veins and thence to the left atrium. Blood proceeding in this direction opens the mitral valve and hence passes to the left ventricle. This chamber of the heart does more work than the others and so the heart muscle (myocardium) that surrounds it is correspondingly thicker. During left ventricular systole the rising blood pressure causes the mitral valve to slam shut and so the only way out is through the aortic valve, thence to the aorta, which is an enormous artery that feeds the other arteries of the systemic circulation.

Each of the four valves of the heart is formed from three leaflets, except for the mitral valve, which has two (and so it also known as the bicuspid valve).

Heart valve disease may be congenital or it may be acquired. If acquired, it is often the mitral or aortic valve that will be affected. For present purposes the most common afflictions are stenosis (where the valve fails to open fully, a common cause being degenerative calcification) and regurgitation (where the valve does not close tightly, thus allowing some of the blood to leak backwards). Regurgitation is also called insufficiency.

Blood is supplied to heart muscle itself by the coronary arteries. They are connected to the aorta, the biggest artery in the body – maybe 25 mm diameter at its root. Here is a simplified diagram. It can be seen that the ascending aorta leads to its arch, in the region of which there are various branches that deliver arterial blood to the upper body. The descending aorta delivers arterial blood to the lower body through various branches, not shown. Two of these are the iliac arteries (10 mm diameter) which supply the pelvis and lower limbs (where they become the femoral arteries, diameter 6-8 mm).

Going back to the ascending aorta, an important thing to notice is the two coronary arteries. They are the sole source of supply of the blood that is required by the muscle tissue of the heart itself (myocardium). It follows that blockage of the coronary arteries, were it to occur, would be disastrous.

It will be convenient at this point to highlight the operational difficulties that would attend the implementation of the invention in suit in clinical practice and to contrast those with mere constructional difficulties i.e. how you build the device that is to be implanted. The patent does not purport to monopolise the operational procedure – in Europe that is not allowed anyway – but only certain apparatus that is to be implanted by that procedure. The operational difficulties loomed large in the evidence in this case.

It will be apparent that the remote implantation into the heart of an artificial valve-on-a-stent by catheterisation could be a cardiological procedure attended by much difficulty and risk. For instance, suppose it was desired to implant the device in the native or usual position at the end of the aorta. Unless the cardiologist had great skill and considerable experience there would be a grave danger of occluding the coronary arteries with fatal results, and especially so if the device failed to remain in place. For one thing, it would be difficult to visualise the movement of the device with the X-ray apparatus commonly available in 1990. Or suppose it was the mitral valve that was defective: how would the catheter safely and reliably be navigated to that site from a blood vessel in the patient’s leg?

Professor Rothman (who gave expert evidence for CoreValve) described a particularly heroic procedure that few cardiologists would care to try – nor their patients, either. This involves the antegrade approach, which was pioneered by Cribier. In essence, access to the heart is achieved by means of an ‘antegrade’ approach – with, not against – the direction of blood flow. The catheter is introduced into a peripheral vein and then advanced along the vena cava to the right side of the heart. If access to the left side of the heart is required then the catheter must be fed through the septum (which is punctured for the purpose) from the right atrium to the left atrium. To illustrate this Professor Rothman, who enjoyed making our flesh creep, produced an enormous needle – it looked about five foot long – and vividly described how it would be used in the procedure just described.

I will say at once that, if it is the law that the patent in suit had to teach the ordinary cardiologist in 1990 not only how to construct a device as claimed therein but how to implant it successfully and safely in a patient, the patent would fail. However, I hold that this was not a requirement of the law and that it would be enough that persons skilled in the art could actually construct such a device. I shall revert to this under the heading of Insufficiency.

In 1990 it was possible to replace a defective heart valve but it required major chest surgery. The patient’s chest was opened, the defective valve was excised, and an artificial valve was sewn in its place. The operation was complex and is described in detail in the expert report of Professor Pepper, an eminent cardiac surgeon who gave uncontroversial evidence in this case. Amongst other things, it required a general anaesthetic and the use of a heart-lung machine. The operation took 3 or 4 hours. The operation was traumatic and hence could not be used for a patient who was too frail, for it might kill him. That patient had to tolerate an inferior quality of life, assisted by such palliative medicine as was available. The procedure is still used today.

The above-described operation was performed by a cardiac surgeon. This professional should not be confused with an interventional cardiologist, who is a not a surgeon but is a physician who specialises in treating diseases of the heart without major surgery. I was told that in 1990 there was a certain amount of rivalry between these two professions, the former tending to regard the latter as trespassing on their patch. Relations were sometimes terse. Nowadays there are many procedures that are performed by cardiologists where, formerly, cardiac surgery would have been required.

In 1990 interventional cardiology was not as specialised a profession as it is now. I shall describe some of the things that had been done.

In balloon angioplasty a diseased blood vessel was expanded mechanically by inflating a balloon that was delivered to the affected site on a catheter. The technique was first demonstrated in 1977 by Andreas Gruentzig of Zurich, and by 1990 it was common to use it for treating a defective coronary artery. The procedure could be carried out while the patient was awake.

A flexible plastic tube called a guide catheter was introduced into a remote blood vessel, typically a femoral artery in the patient’s groin. It could be introduced through a needle (rather like injection) or by making a small incision called a cut-down. (This is called percutaneous access.) A fluid opaque to X-rays was introduced and so the cardiologist could visualise his apparatus and steer it up the artery to the correct site. A fine guidewire was inserted through the guide catheter and into the coronary artery. A special balloon was mounted on the guidewire and conveyed to the affected site. The balloon was then inflated so as to dilate the coronary artery, thus curing the problem. That was the hope, anyway, for the technique did have its problems.

When this technique was introduced into hospitals it was necessary for the cardiac surgeon to stand by in case emergency heart surgery became necessary. For example, the coronary artery might suddenly close again (restenosis).

One technique for preventing restenosis was to use a stent. As indicated, a stent is a supporting scaffold, typically a tube, made of wire, perforated metal sheet or other suitable material. The stent could be mounted on the balloon and expanded to hold the blood vessel open.

A variety of stents were introduced and some were self-expanding. Early stents were made of bare metal but nowadays they can be impregnated with drugs: the idea is to prevent the formation of blood clots.

Sometimes a technique called valvuloplasty was tried. In this, a balloon was delivered to the interior of an aortic valve that had become stiff through calcification and expanded in the hope of improving its patency (openness). The idea was to make the patient’s natural heart valve behave better.

Stents were also employed by cardiac surgeons in order to hold artificial heart valves in place. Of course in this case their use required open chest surgery. A variety of such stents were designed and used.

By 1990 stents were well known to cardiologists, but their use was limited. Perhaps the best known was due to Palmaz. Schools of thought differed: some thought stents were the way of the future, others that little would come of them, still others that their use was for emergency surgery only.

It appears that the invention of the patent in suit was largely the work of Dr Henning Andersen, a Danish doctor. In practical terms Dr Andersen took the work no further forward than certain experiments in pigs. These were not particularly successful but the work was published in professional refereed journals. In passing, the reaction of the referees was that the idea was bold and interesting but somewhat too ambitious. It also appears that Dr Andersen did not succeed in persuading clinicians to take up his idea and he eventually sold his patent; in the end it was acquired by Edwards. In my judgment those facts, in themselves, are neutral to what I have to decide. They are consistent with what is a common situation in patent practice: a one-man band failing because large resources were necessary in order to achieve clinical validation in what was a very risky branch of medicine.

The patent in suit claims priority from a Danish application and it has suffered slightly in translation. There are few practical instructions and those that are there are concerned with a prototype device used in experiments with pigs.

As illustrated in Figure 2, the patent describes a stent made of two folded surgical stainless steel wires 2, 3 to form two superposed rings secured together by sutures and having three tall loops 4. To this stent a natural pig heart valve 6, suitably cleaned, is secured by suturing. (I understand that Dr Andersen obtained his pig valves from a butcher.) The commissural points 5 of this heart valve are secured to the tallest loops, thus providing a sort of “three-cornered hat” structure. “Commissural points” is something of a patent agent’s expression but it caused no difficulty to the experts in this case. Essentially, the commissural lines of a heart valve are where the edges of the valve leaflets meet, and so a commissural point is where the corners do. The patent indicates that a suitably adapted valve would be used in the case of humans and it says that instead of a biological valve it would be possible to have a valve made from synthetic materials.

The patent also describes a device with a stent with a closed wall, which may be self-expanding once in place.

All of the devices shown in the diagrams of the patent have stents of essentially cylindrical conformation.

Although the patent refers to trials on pigs, it contains no results of those trials.

Claim 1 of the patent reads as follows:

The following witnesses testified for CoreValve: Professor Martin Rothman (an eminent cardiologist); Dr Richard Hillstead (who had experience in stent design); and Mr Robrecht Michiels (CoreValve’s President and CEO).

The following witnesses testified for Edwards: Dr Nigel Buller (an eminent cardiologist); Professor John Pepper (an eminent heart surgeon); Dr Anthony Lunn (who had experience in stent manufacture); and Mr Stanton Rowe (Edwards’ Vice President of Advanced Technology).

I thought all of these men were honest witnesses. Those who were expert witnesses impressed me with their knowledge of their subject matter. The witnesses of fact were generally reliable, although of course there were questions of degree of emphasis.

Both sides mounted the usual attacks on the other side’s experts: they were said not to be impartial, and so on. Beyond the obvious and inevitable fact that every man in the world has his own personal eccentricities of approach, I reject those attacks. They were probably made because each side worried the other side would do the same. Counsel fear too much that courts in patent cases will be unduly influenced by the opinions (i.e. value judgments) of experts, as opposed to the cogency of their reasoning.

In Alan Nuttalt Ltd v. Fri-Jado UK Ltd [2008] EWCH 1311 (Pat) I said, largely quoting from Jacob LJ:

In weighing the evidence in this case I have made allowances for the personal attributes and prejudices which these witnesses - like all of us - inevitably have.

So in the present case. Thus I have allowed for the fact that, for example, Professor Rothman is an enthusiast by temperament, very alive to the latest innovations in his profession; and for the fact that Dr Buller has been (as they say in theatrical circles) “on the road” for a long time now, having given expert evidence in many patent cases. Both witnesses were excellent.

The first words of Claim 1 of the patent are ‘A valve prosthesis, preferably a cardiac valve prosthesis, for implantation in the body’. The word “preferably” imposes no limitation; but it does indicate that, whatever other type of valve prosthesis is covered by this claim, it does at least cover a cardiac valve prosthesis. (Footnote: ¹) The word “for” at the start of a patent claim is conventionally taken to mean that the device is suitable for the stated purpose, and not that it is necessarily intended for that purpose, still less that it is actually used for it. For example, if a piece of prior art is as a practical fact capable of being used for a certain purpose, it may anticipate a patent claim even though it would never occur to anyone so to use it. If a patentee wishes to protect himself from that kind of attack his proper course is to limit his claim accordingly. The classic illustration is Adhesive Drive Mounting v. Trapp (1910) 27 RPC 341 where a claim to a sheet that would become tacky when heated “for” sticking photographs in albums was held to be anticipated by a sheet (called a “pellicle”) that did have those physical properties even though nobody had previously suggested its use for that special purpose. The inventor of the pellicle had anticipated the later claim as M. Jourdain had spoken prose – without knowing it.

Hence the claim in the present case covers a cardiac valve prosthesis that is in fact practically suitable for implantation in the human body, whether one knows it or not. I do not understand those propositions to be disputed.

The following words and phrases in Claim 1 of the patent were in contention.

‘Elastical’. This is a rather poor translation, but I make the necessary accommodation and read it as ‘elastic’. According to Edwards it means that the stent and the valve will deform when loaded and recover when the load is removed e.g. according to the contractions of the heart. According to CoreValve it imports no real limitation at all because the Claim anyway informs us that the stent is radially collapsible and re-expandable.

The word occurs in a context of a stent and valve that are stated to be elastic and radially collapsible and re-expandable during implantation by means of catheterisation. “Elastical” should therefore mean something more than merely being collapsible and re-expandable. It suggests a certain inherent springiness. This is confirmed by the description in the patent. It tells us that the device of Fig 2 is made of stainless steel wire. At column 3 lines 10-20 the patent says:

At column 7 of the patent there is a description of implanting the valve prosthesis by means of a balloon catheter. At lines 40 to 45 it is stated that the outer diameter of the device when fitted on the balloon is 10 mm but after inflation of the balloon it achieves a diameter of 30 mm. Then at line 52 it says: “The balloon catheter 11 may subsequently be removed from the aorta 10 (Fig. 7). Due to the stiffness of the metal the valve prosthesis will prevent a contraction. However, smaller contractions may occur (<10% diameter reduction) after the deflation of the balloon catheter 13.” This suggests that the device is not rigid but is elastic under hemodynamic loading and unloading, but not necessarily to the extent of being self-expanding on implantation i.e. without a balloon catheter to assist (Footnote: ²).

In conclusion, therefore, I hold that “elastical” refers to the springiness of the stent and valve i.e. they have some capacity to flex and recover their original configuration at least when a hemodymanic load is applied and removed, but not necessarily so when the load is heavy i.e. during insertion. A device devoid of flexibility is excluded.

‘Cylindrical’ and ‘cylinder surface’. Claim 1 requires that the stent is made from a “cylindrical support means” and that the commissural points of the valve are mounted on the “cylinder surface” of that stent. The meaning of those phrases is highly in contention. The reason for that, as we shall see, is that CoreValve’s device has not a uniform cross-section, but is bulbous at one end.

According to Edwards, there are four possible conditions in which to consider the stent:

in its condition as manufactured i.e. before being assembled into a catheter;

in its condition when ready to be introduced;

in its condition once implanted in the body (when it will tend to conform to the shape of the site);

in the case of a self-expanding stent, in its condition when “free in air”.

Edwards rejects conditions (a), (c) and (d) and urges condition (b): more precisely, they contend that the device is collapsible into a cylinder so that it can be delivered by catheter. They say that a third party should be able to know whether their product infringes without having to know its conformation when inside the human body; and that the “free in air” approach cannot be applied consistently to self-expanding and balloon-expandable stents.

CoreValve say that the device must be cylindrical with a constant cross-section when the device is free in air i.e. before being compressed onto the balloon or delivery catheter.

I hold that Claim 1 of the patent in suit requires that the stent be generally cylindrical (i.e. substantially cylindrical but not mathematically so) in its natural and free condition as manufactured, and not in a condition where it is deformed whether by being compressed onto a balloon or inserted into a patient. The Claim says that the stent is “made from a radially collapsible and re-expandable cylindrical support means”. The Claim is addressing a valve prosthesis “for” implantation in the body, not one that is already there. All of the Figures in the patent show a generally cylindrical stent.

Mr Wyand QC, who appeared for Edwards, relied on the following passage which occurs in column 8 of the patent:

But, said Mr Wyand, the cross-section of that part of the aorta is not uniform. Therefore, said Mr Wyand, the description in the patent implies that the cross-section of the stent need not be uniform. In my judgment, that is to place more inferential weight on the above-quoted passage than it can bear, the more so since it is not all that clear in the first place. In my judgment it probably means no more than this, that a taller stent can be used.

‘Radially collapsible and expandable’. This means: for the purpose of delivery and implantation by catheter – not just in the abstract.

‘By a technique of catheterization’. Claim 1 of the patent requires that the stent be made from a radially collapsible and re-expandable cylindrical support means for folding and expanding together with the collapsible valve “for implantation in the body by means of a technique of catheterization”. The controversy here is that CoreValve contends that the device must be capable of being introduced percutaneously by puncture or a small cut-down, under a local anaesthetic; but Edwards contends that it means no more than this, that the device can be introduced by a catheter without having to open the thoracic cavity.

I hold that Edwards is right on this. The Claim is not making any refined point. It is clear to me from a reading of the antecedent description as a whole that the purpose is to have a device that can be implanted without the previously existing need for major, open chest surgery – which required the use of a heart-lung machine and general anaesthesia. As Mr Wyand QC rightly observed, implantation by a technique of catheterisation need not imply percutaneous delivery, which is but a sub-species thereof.

In my judgment, the device supplied by CoreValve has all of the attributes called for by Claim 1 of the patent in suit except that its stent is not “made from a … cylindrical support means”, nor are the commissural points of its valve mounted on the “cylinder surface” thereof. Instead of having a stent with generally cylindrical shape i.e. with a substantially constant diameter, the CoreValve product is bulbous at its end remote from its valve.

#

#

This is said to have the advantage of better securing the device in the aorta. Be that as it may, it is certainly not the case that the stent is generally cylindrical. The above drawing understates the position, because the CoreValve device as supplied – in a jar, as it were – is not deformed inwardly by the aorta as shown in the drawing, and so its diameter at its bulbous end is about twice as great as at its narrow end.

Mr Wyand had an argument to the effect that the lower portion of the device is cylindrical and so it should be considered as a cylinder with another cylinder on top. It made no impression on me. Any geometrical figure can be considered to be made out of others. At that rate, a square could be described as “triangular” because any square comprises two triangles. And there is no doubt that the commissural points of the CoreValve device are mounted on the bulbous part.

I hold, therefore, that the CoreValve product does not infringe the patent in suit.

Section 60(5)(b) of the Patents Act 1977 provides that an act which would otherwise infringe a patent shall not do so if “it is done for experimental purposes relating to the subject-matter of the invention”.

This point is somewhat hypothetical, in that I have held that CoreValve are not using the invention. If they were, they would be supplying a product different from what they are using in reality. It would be cylindrical instead of bulbous. The statutory exception does not permit a patented invention to be used for experimental acts relating to a different invention.

I can state the relevant facts briefly. At present CoreValve does not supply its product to all comers, but only to selected hospital sites in Europe (not the USA). This is part of a formidable clinical programme, referred to in the evidence as the Registry, an important feature of which is that cardiologists are carefully trained in the use of the product, and then train others – a sort of pyramid apprenticeship scheme, as it were. As will become more apparent later, it is highly important that there should be professional confidence in these products if they are to take off commercially. This scheme helps to establish that confidence.

CoreValve obtained a CE Mark [i.e. regulatory] approval for its product in May 2007 for the treatment of aortic stenosis in elderly high risk patients. The current product is Generation 3 so far as physical design is concerned. However, and according to its CEO Mr Michiels:

Apart from the reference to decisions on future design generations, which is rather vague, and would have applied to Henry Ford when he first introduced his Model T automobile, all of these investigations are about cardiological procedures, and do not concern trying modifications to the hardware as such.

CoreValve do not supply their product gratis; on the contrary, they invoice a very substantial amount for each unit.

It is well settled (Monsanto v. Stauffer [1985] RPC 515, C.A.) that mere field trials which are intended to demonstrate the efficacy of the product for the purposes of regulatory approval do not qualify for the exception set for in s.60(5)(b) of the Act. In general, the purpose of this defence is to encourage scientific research while protecting the legitimate interests of the patentee. This involves a balance.

Section 60(5)(b) is based on Article 27(b) of the Community Patent Convention. The Federal Supreme Court of Germany considered the equivalent provision in Klinische Versuche (Clinical Trials) I[1997] RPC 623. The only part of the court’s official headnote that is relevant for present purposes is as follows (English translation):

In that case the substance in question (an interferon) was known for use in the treatment of rheumatoid arthritis and the defendants were conducting clinical trials to see if that substance could be used for treating other diseases such as cancer, AIDS and hepatitis. The invention – the thing that was claimed in the patent – was the substance as such. I can see that those clinical trials were squarely within the purpose of the exception, for their immediate purpose was to generate scientific information by experimenting with the substance that was the subject of the patent claim.

However, there must surely be an outward limit to that principle. Suppose the defendants in the German case had been selling a pharmaceutical that was fairly new to the market and their defence had been that, by so doing, they were gaining valuable information that was not otherwise available – contraindications, for instance, which could be stated in the product literature. Would that be acts done for ‘experimental’ purposes?

A defendant could always say, and with some truth, that by putting his product on the market (general or special) he was gaining valuable information that might even prompt him to modify his device in future. I have referred to Henry Ford’s Model T car. I dare say that vehicle went through various modifications in the light of experience on the roads of early twentieth century America, and that is usually the case with any engineering product.

I acknowledge that the mere fact that the purpose of the defendant is commercial is no rebuttal of the statutory defence. After all, most pharmaceutical research organisations are commercial. They do research because they hope to make money one day. However, in the present case it cannot be denied that an immediate and present purpose of CoreValve is to generate revenue – which was not so in the German case.

I therefore think that a more complete statement of the principle – it did not arise in the German case – should involve the consideration whether the immediate purpose of the transaction in question is to generate revenue.

The relevant statutory phrase is “acts done for experimental purposes”. The difficulty arises where the defendant has mixed purposes. I would reject the extreme proposition that, so long as one of the defendant’s purposes is to generate information of scientific or technical value, it is irrelevant that another of his purposes is to generate ready cash. There may be no help for it but to consider the defendant’s preponderant purposes.

On the evidence in this case I would hold that CoreValve’s purposes are threefold: (1) to establish confidence in their product within the relevant market; (2) to generate immediate revenue of a substantial character; and (3) to gain information about clinical indications and, possibly, future modifications to be made to the physical structure of the device in the light of experience. I do not find that purpose (3) was their preponderant purpose.

I have not found this point easy, but on the whole I would hold that, on the assumption that the CoreValve device falls within Claim 1 of the patent in suit, section 60(5)(b) of the Patents Act 1977 is not a valid defence on the facts of this case.

Section 72(1)(c) of the Act provides that a patent may be revoked if its specification does not disclose the invention clearly enough and completely enough for it to be performed by a person skilled in the art. Mr Watson QC, who appeared for CoreValve, mounted a formidable attack on the sufficiency of the patent in suit.

Earlier in this judgment I have outlined some of the major procedural or operational difficulties and dangers attendant on successfully and reliably implanting a device of this character in the human heart. I now add that nobody implants an artificial heart valve in humans in any circumstances save in cases of necessity. I shall amplify on that by way of illustration. If you were told that you were certain to die of heart disease within a twelvemonth unless you had an artificial valve implant soon, but that the operation had a 90% chance of killing you at once, what would your rational choice be? In truth there could be no rational choice or calculus of risks and benefits. Some might prefer the chance of possibly living and seeing their family and friends for a year. In any case clinical prognostications could seldom be so precise. In short then, it is a natural human reaction to be cautious, and that is bound to be reflected in the attitudes of the clinical professionals who are human too and care about their patients.

In extreme circumstances – say, if you were going to die within a fortnight anyway – the rational choice would be to go for the operation. However, in the case of a novel and risky operation it is seldom possible to have much confidence in its likely clinical outcome: it is not a question of percentages at that stage. For a reasonably confident prognosis to be offered it is necessary to build up a body of clinical experience. And in order to build up that experience it is necessary to proceed cautiously and in small stages, starting with patients in extremis: elderly or frail patients who could not run the risk of open chest surgery and whose heart valve disease presented a high risk of morbidity.

That would be preceded with extensive experimentation in animals. For present purposes no available experimental animal is really equivalent to man. For example Dr Andersen experimented on pigs but they have narrow femoral blood vessels. In practice experimentation on animals on the necessary scale would have to be undertaken by an organisation with substantial resources.

One line of Mr Watson’s attack was that it took Edwards and their predecessors a long time to come up with a product that could be entered in clinical trials. It would not be appropriate to reproduce all the details in the body of this judgement, and so I have set out CoreValve’s chronology in an Annex. I have already said that I do not regard early experimentation by Andersen and others to be very determinative because I believe that a product of this sort stood no chance unless it was backed by the resources of a large organisation. Now, Mr Watson’s point might have been a telling one if it had been shown that Edwards had to keep making modifications to their product to get it up to standard – implying an unacceptable degree of trial and error, albeit routine – or that it required a brainwave in order to do it. But it has not been shown to my satisfaction that anything of the sort happened. The facts are consistent with what happens quite often in cases about patents in the medical field.

Mr Watson contended that the patent simply does not describe a device that is small enough to be implanted as the Claim requires. However, and despite the evidence of Dr Hillstead, I find that it would be not beyond the capability of a manufacturer of stents to fabricate a device that was small enough, given a reasonable amount of trial and error.

Therefore, I do not hold that the patent’s description of the device itself was unduly wanting. True, Dr Andersen described a prototype device for testing on pigs. But the patent indicates that appropriate dimensional modifications would be required for implantation in humans, and I do not find it to be lacking in that respect.

Another line of attack was that it is not enough if the patent adequately describes a physical device. It was necessary (it was said) for it to go on and describe how that device was to be implanted clinically. Professor Rothman’s evidence demonstrated, to my satisfaction at least, that there would be formidable problems. For example, the device would have to be implanted so that it did not dislodge and block the coronary arteries. Just to navigate the device on a guidewire using the clinical X-ray apparatus that was generally available would be exceedingly difficult.

I believe that Mr Watson is right to say that the patent does not contain sufficient directions judged by that standard. Indeed I doubt that any patent could be sufficient in that respect, because it would be akin to learning to fly a helicopter by reading the manual. In our case there could be no substitute for clinical hands-on experience.

I have referred to aircraft. The real merit of the Wright Brothers was perhaps not so much that they invented a device that was capable of controlled heavier-than-air powered flight, but that they realised that it would be a strong idea to teach themselves to fly before actually flying. And so they proceeded in small, cautious steps, starting with gliders. Their first flight in their powered machine covered a distance that was less than the length of the aisle in a modern Boeing 747. It took several years before any realistic distances were achieved. But nobody says that the Wright patent should have been held invalid for lack of enablement.

I stress those points because, in my judgement, CoreValve’s case confuses the sufficiency of the instructions for making a device falling within the claims of the patent and the sufficiency of instructions for telling the reader how to use it.

There was much evidence about the difficulties of implanting the device in the iliac artery without needing a general anaesthetic. But the patent was not required to teach the reader how to do this. In my judgment it was enough that the skilled device manufacturer could, by reasonable trial and error, construct a device that could be inserted via the femoral route. I do not believe the patent had to go on and describe all possible operational alternatives.

After the conclusion of the argument and while I was considering this judgment I was troubled by the point I have mentioned in paragraph 3 above. In outline, it is this. On obviousness over the prior art (see below) is it not relevant that the person skilled in the art would be put off considering any modifications to a prior art device that might lead him within the ambit of Claim 1 of the patent in suit because he would sense that implantation of the device in clinical practice would be both difficult and dangerous – lions in his path? If so, is it not something of a double standard to hold that, on insufficiency, it is enough if the patent gives directions which would enable the skilled person to get there clinically in the end, by cautious trial and error?

On consideration, I do not think so. I believe that the point goes, not to the validity of the patent, but to the relief that would be granted had it been shown that the patent was infringed. In the event the point is not before me. Therefore I do not have to decide whether a court, in the exercise of its discretion concerning what is an equitable remedy would refuse an injunction that would prevent the supply of life-saving medical devices, not otherwise available, and relegate the patentee to a reasonable remuneration.

I say this because in the world of patents for medical products there is often a gap between aspiration and achievement. Nobody suggested that the patent that covered cimetidine was invalid on the ground that it disclosed that substance and innumerable others too, without stating which was the one to take to clinical trials.

A still further line of attack was that the patent in suit does not disclose any actual results – even in pigs – that demonstrated that the device would work. That argument was based on the decision of the Court of Appeal in Conor Medsystems Inc v. Angiotech Pharmaceuticals Inc, although strictly speaking that was a case about obviousness where the patent did not contain results that demonstrated that the thing that it claimed would work.Mr Watson put his case on both grounds and it will be convenient to deal with it here. The Angiotech decision has now been reversed by the House of Lords: [2008] UKHL 49. At paragraph 37 Lord Hoffmann said:

In my judgment, the patent in our case does pass the plausibility test. I do not mean that the skilled reader would be confident that the device would work clinically with complete reliability and safety. But, because it was known that stents, as such, were used and used successfully in at least some cases, I do not find that he would think that the idea was implausible as a matter of principle.

US Patent No 3,657,744 was granted to Ersek on 25 April 1972. Therefore for the purposes of the law of anticipation it must be interpreted as of that date.

For present purposes the law of anticipation is too well known to require explication. CoreValve must show that Ersek contains clear and unmistakable directions to make a device that would, in fact, fall within Claim 1 of the patent, whether anyone knew or intended it or not: Synthon BV v. SmithKline Beecham plc[2005] UKHL 59.

Ersek proposes a stent, with valve attached, for the purpose of implantation in a patient who is undergoing conventional open chest surgery. His aim is to have a device that the surgeon can implant rapidly, thus reducing the time required for that operation where “every minute is important”.

In order to achieve that aim Ersek proposes a cylindrical tubular sleeve with valve attached. Prior to surgery this is mounted on a purpose-built expansion tool, which has a sort of pistol grip with a trigger. During the operation the surgeon forces the device into place and expands the sleeve using the tool. The sleeve is an openwork cylinder made of a stainless steel. When it is expanded its sharp edges embed themselves into the wall of the blood vessel. These help to make the device stay in place because a natural biological process occurs (endothelialisation) in which tissue rapidly covers the metal.

There is no suggestion in Ersek that his device should be implanted ‘by a technique of catheterisation’. However that does not matter so far as the law of anticipation is concerned, provided the document contains clear and unmistakeable directions to make a device which could be implanted by catheterisation.

I do not find any clear and unmistakeable directions in Ersek to make the sleeve elastic in the sense that I have discussed above on the construction of the patent in suit. It depends on the grade of steel, the precise configuration and dimensions of the sleeve, and so on, as to which this patent says practically nothing, except that we are told that it retains its shape once it has been expanded by the tool by about 50%. That is enough to dispose of the anticipation attack based on this prior art.

I also do not accept that Ersek contains clear and unmistakeable directions to make a sleeve that is “radially collapsible and re-expandable” in the sense of Claim 1 of the patent in suit i.e. for implantation in the body by means of a technique of catheterisation. Because the device called for by the Claim is to be delivered through a narrow passage, and not implanted in open heart surgery as in Ersek, I hold that there are no clear and unmistakeable directions to make a sleeve that can, in fact, be collapsed down so as to fit within the tube of the catheter system and retain those metallurgical properties which will enable it to be re-expandable e.g. by a balloon. What about the elastic limit of the material?

Here the question is whether it would be obvious to the skilled person – who may be a team of persons each with appropriate skills – in 1990 to modify the Ersek device so as to arrive at a device which would answer to the requirements of Claim 1 of the patent. It is elementary that whatever the skilled readers do, they do it without hindsight.

The structured approach to considering obviousness was recently re-stated by Jacob LJ in Pozzoli Spa v. BDMO SA [2007] EWCA Civ 588 at §23:

I answer those questions as follows:

As to the last point, a team that would make a sleeve with the requisite properties would have to have a reason for doing so. Unless they did it by pure accident, and I am not persuaded that they would, they would have to think: “Forget about that pistol device, open chest surgery, and so on. Put it up a blood vessel in the patient’s leg”. Stated that way, it is blatant hindsight.

A more persuasive argument is that, upon perusal of Ersek, the cardiologist would say: “That reminds me of the Palmaz stent. I know those are implanted by a technique of catheterisation, so why not construct an Ersek device that could be implanted that way? I had better go and see a stent fabrication specialist”.

Professor Rothman thought that that would have been his reaction. But he did not in fact read Ersek before the priority date of the patent in suit, and it is no disrespect to him that this is necessarily speculation. Besides, I am far from persuaded that Professor Rothman corresponds to the unimaginative skilled man who is the touchstone of patent law. Quite the contrary.

This argument seems persuasive only because we already know about Andersen’s invention. I believe that by 1990 the interventional cardiologist upon reading Ersek would find it deeply unattractive, and put it aside. He would think (but not be rude enough to say) “That is out of the dark ages, and a job for surgeons”.

US Patent No 4,922, 905 was granted to Strecker on 8^th May 1990, that is to say, 10 days before the priority date of the patent in suit.

Stecker’s general idea is to have a dilatation catheter for treating e.g. arterial walls, so devised “as to preclude as far as possible any collapse of the vessel walls”. There is a stent which is radially expandable by the balloon and which, once expanded, stays that way.

I shall not waste many words on Strecker. It is really cited because at Figures 12-14 there is described an internal aortic valve 51. The device comprises a tubular section 52 “made by crocheting, knitting or the like” and a valve 51 consisting “for example, of a finely meshed web of woven wires whose wires crossing each other run in parallel with two rhombus (diamond) edges”. Figure 12 (on the left-hand side, below) shows the device before radial dilation and Figure 12 (right-hand side) after radial dilation; in these Figures only one leaflet is visible.

Figure 14 below may give a better idea, or it may make confusion worse confounded. It is by no means clear how this valve arrangement is supposed to work and in my judgment is was not clear to the experts in this case either.

I hold that Strecker does not enable – teach the skilled reader how to make – a device that answers to the requirements of Claim 1 of the patent in suit. I also hold that the stent is not “collapsible and re-expandable… for implantation in the body by means of a technique of catherization”. In Strecker the stent is not collapsed, but starts off with the desired size for introduction on a catheter.

CoreValve’s case is really that “If the skilled person did not wish to pursue the precise design of valve leaflet set out in Strecker, it would be obvious in 1990 to replace it with a known and available biological or pericardial alternative”. In my judgment that is sheer hindsight.

The patent is valid, but not infringed. I must dismiss the claim and the counterclaim. If either side wishes to apply for permission to appeal, I would be minded to grant their request.