England and Wales High Court (Patents Court) Decisions

IN THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
PATENTS COURT

		8th November 2000

Mr Justice Laddie:

Introduction

1. By this action, Lilly ICOS LLC ('Lilly') seeks to invalidate and revoke European Patent (UK) 702,555 ('the Patent') which is registered in the name of Pfizer Limited. The patent seeks protection for the medical use of a number of chemicals in the treatment of impotence. One of these has the chemical name, sildenafil citrate. It is the active ingredient in the anti-impotence drug known as Viagra. Indeed it is part of Pfizer's case that this is the patent 'for' Viagra. It argues that the commercial success of Viagra is testimony to its inventiveness. Lilly is in the course of producing its own anti-impotence drug and it fears that the patent will stand in its way. It says that the patent is invalid because it is anticipated by a single piece of prior art, obvious over a number of pieces of prior art, insufficient and contains added matter. Before considering the patent and the attacks levelled at it, it will be useful to have a basic understanding of the anatomical, physiological and chemical background against which the issues have to be decided.

Technical Background

2. The following technical background is derived in part from the primer agreed by the parties and supplied to me before the trial and partly from the written reports, witness statements and oral testimony given by the witnesses at the trial. I believe it to be uncontroversial. It represents part of the common general knowledge at the priority date of the patent in June 1993.
3. The penis, like any other living organ or tissue in a mammal, is supplied with nutrient-rich and oxygen-rich blood in vessels called arteries. The flow is maintained by allowing the exit of the blood from the organ through vessels called veins. In an adult male human the penis has to grow in length and become rigid in order to facilitate sexual intercourse. This is called erection, or tumescence. After sexual intercourse or removal of the sexual stimulation which led to tumescence, in normal males the penis will return to its more flaccid and shorter resting state. This is called detumescence. Tumescence and detumescence are caused by changes in the blood flow inside the penis. To understand how this is achieved, it is necessary to know something of the basic anatomy of the normal penis.

(a) Anatomy.

4. Set out below is a diagrammatic representation of a cross-section of the penis^[1].

5. Towards the bottom of the figure the urethra is identified. This is the channel through which urine and semen can travel. Above and to each side of the urethra are two symmetrical compartments, or corpora, each of which is called a corpus cavernosum. It is changes in the blood flow inside these compartments which lead to tumescence and detumescence. The corpora cavernosa consist of vascular sinusoids. That is to say they contain a network of very small blood vessels or passages. The sinusoids are surrounded by a type of muscle known as cavernous smooth muscle which, like any other muscle, can contract or relax. The corpora cavernosa also contain a matrix of supporting connective tissue. Blood is supplied to the vascular sinusoids in the copora cavernosa through a network of arteries. The blood drains out again through a network of small veins or venules which drain into larger veins, called emissary veins. The latter veins lie on the outer surface of each corpus cavernosum. Surrounding the corpora cavernosa is a fibrous layer called the tunica albuginea. Rather like a sausage skin, it is not very elastic. The emissary veins are located between the corpora cavernosa and their respective tunica albugineas.

(b) function

6. When the penis is in its flaccid resting state, the flow of blood into the corpora cavernosa is restricted or throttled back by constriction of the smooth muscle which surrounds vessels in the incoming arterial network. These muscles act rather like ligatures. When they contract they reduce the internal diameter of the vessels in the arterial network. Of course they do not completely close the arteries or the penis would be completely starved of blood. When the smooth muscles are contracted like this, less blood can flow into the corpora cavernosa. Similarly the cavernous smooth muscle is contracted to the same effect. On the other hand the venules and veins are unaffected. They continue to be able to remove blood from the penis with comparative ease.
7. When an erection is triggered, the smooth muscles surrounding the vessels in the arterial network and the cavernous muscles relax. The arteries open up. It is now easier for them to supply blood to the corpora cavernosa. Blood floods in to the sinusoids. The sinusoids start to swell and each corpus cavernosum presses up against the surrounding tunica albuginea. The swelling of the sinusoids squeezes the venules, thereby reducing the size of their internal passages. This reduces their ability to drain blood from the corpora cavernosa. Similarly, as the corpora cavernosa expand and press against the tunica albuginea, the emissary veins located between these two tissues are squeezed and are less able to drain blood from the penis. The result is that the penis becomes engorged with blood and stiff. A state of full erection is achieved when the pressure in the corpora cavernosa equals mean systolic pressure (i.e. the pressure of the blood leaving the heart). When the erectile process works in reverse the smooth muscles contract and the arteries again become constricted. This reduces influx of blood. At the same time the smooth muscle of the sinusoids become constricted which, in turn, reduces the external pressure on the venules and the emissary veins and allows the blood to flow out. So, during an erection inflow of blood is facilitated and outflow hindered and during detumescence outflow is facilitated and inflow is hindered.
8. It should be noticed that contraction of the smooth muscles results in relaxation or detumescence. On the other hand relaxation of the smooth muscles results in tumescence/erection. This is because the smooth muscle is being used to throttle back the blood which, if allowed to flow in at arterial pressure, would result in the penis being gorged with blood and erect.

(c) Smooth muscles

9. The body of a mammal contains a number of different types of muscle. For example the muscles which result in one being able to raise or lower an arm are part of the 'voluntary' system of the body and are under conscious control of the brain. On the other hand there are muscles in the body which are under 'involuntary' control. These muscles are made to contract or allowed to relax by the background housekeeping systems which exist in mammals (and other animals). The smooth muscles which surround the sinusoids and small arteries in the corpora cavernosa are part of this 'involuntary' system. They are supplied and brought into operation by a separate system of nerves to those used in the voluntary system. Various types of smooth muscle are to be found in different tissues and organs of mammals.

(d) Male Erectile Dysfunction (MED) and its treatment.

10. A significant part of the adult male population suffers from male erectile dysfunction. An accurate figure for the numbers affected is not known, partly because some sufferers are reluctant to acknowledge the existence of the problem or to discuss it with third parties. However the sufferers of one form or another of MED in Britain may be in the millions. MED becomes more prevalent in older men. In some males MED takes the form of a total inability to achieve an erection. In others the erection may be incomplete or last insufficiently long to achieve any or satisfactory sexual intercourse. In others the penis may stay erect for a long time or permanently (priapism). It is probable that there are a number of different causes of MED, with some males having one defect and others having others.
11. There were, at June 1993, a number of well know treatments for men who were unable to achieve satisfactory erections. The most widely used consisted of the self administration, by injection directly into the corpora cavernosa, of various drugs. The injection had to be effected shortly before sexual intercourse was desired. The drugs deployed included phenoxybenzamine, phentolamine, papaverine and prostaglandin E₁. In England papaverine was the most widely used of these agents. For obvious reasons this form of treatment did not suit everyone. Some males disliked the act of self injection and sometimes one or both partners found the treatment discouraging. Furthermore the injections were not without side effects. For example frequent injections directly into the penis cause scarring. Sometimes treatment caused pain or priapism. This form of administration of the drug is frequently referred to as intracavernosal or 'i.c.' As an alternative to i.c. treatment, some drugs were injected into the urethra. Again there were side effects.
12. Another type of treatment consisted of the oral administration of a drug known as yohimbine. Although it was accepted before me by all the witnesses that the oral administration of a drug was the avenue of choice, there was much doubt as to the efficacy of yohimbine. As a result, i.c. injections were used widely not withstanding their obvious disadvantages. A number of other treatments were used in a minority of cases. These included the use of suction devices and prostheses, the use of glyceryl trinitrate patches applied to the penis and sometimes increases in arterial supply to the penis effected by surgical intervention. All suffered from some side effects even in those cases where they were effective in curing or reducing MED.

(d) The Chemistry of Smooth Muscle Relaxation

13. At this point it will also be convenient to explain what is going on inside smooth muscle which makes it contract and relax. All the matters set out below were known at the priority date and were common general knowledge.
14. Muscle is made up of living cells. In the case of the type of muscles distributed in the penis, each cell is made to relax when it receives a chemical 'messenger' which triggers the series of chemical reactions in the cell which cause the relaxation. In smooth muscles the creation of such a chemical messenger can be traced back to the production of a short lived gas, called nitric oxide (chemical formula: NO). Nitric oxide is produced or released from at least two sources. First smooth muscle is lined with cells called endothelium cells. It was known in the 1980's that this produced a 'relaxing factor', that is to say something which affected the contraction of the muscle. Because of its source and effect this was known as EDRF (i.e. Endothelium Derived Relaxing Factor). EDRF was discovered to be nitric oxide in the late 1980's.
15. Secondly the nerves which service smooth muscle also release nitric oxide. These nerves are sometimes called non-adrenergic non-cholinergic (to distinguish them from certain other nerves in the body which are adrenergic or cholinergic) or NANC nerves. In both cases the nitric oxide is produced by a reaction from a chemical called L-arginine. The chemical reaction is catalysed by an enzyme called nitric oxide synthase.
16. The nitric oxide produced by these two methods enters the smooth muscle cells and activates another enzyme (called guanylate cyclase) which converts another chemical called guanosine triphosphate (GTP) to cyclic guanosine monophosphate (cGMP). cGMP in turn activates certain other intracellular enzymes (called protein kinases) which cause or promote chemical reactions which relax the smooth muscle. cGMP can itself be inactivated. A group of enzymes, called phosphodiesterases or (PDEs), cause cGMP to be turned into a chemical which is ineffective to relax muscle. This ineffective chemical is known as 5'GMP. Thus turning cGMP into 5'GMP has the effect of removing the agent which causes the muscle to relax. This sequence of chemical reactions can be represented in abbreviated and diagrammatic form as follows.

17. This is known as the NANC pathway. Notice that the agent which is believed to cause relaxation is cGMP. That agent is produced by the 'front end' of the pathway and is dependent, amongst other things, on the generation of NO. It is removed by the 'back end' of the reaction - i.e. the part which is dependent, amongst other things, on the action of the PDE enzymes. As will be recalled, relaxation of the smooth muscle results in erection. So, production of cGMP should cause or contribute towards an erection and removal of it should cause or contribute towards detumescence.
18. What goes on in the penis is more complex than this. Penile smooth muscle is also relaxed by other agents. In particular there is another chemical called cyclic adenosine monophosphate or cAMP which relaxes the smooth muscle like cGMP does. It is produced from a different starting material to cGMP in response to different messengers. In particular the messengers which trigger the creation of cAMP are called vasoactive intestinal peptide (VIP) and prostaglandin E1 ('PGE1'). Like cGMP, cAMP is turned into an ineffective chemical by the action of a PDE enzyme. By the priority date it was well known that there were a number of PDEs. Each one regulates the level of cGMP, cAMP or both to varying degrees. By June 1993, five different PDE enzymes or enzyme groups were recognised by scientists and a classification system was adopted to discriminate between them. By that time it was also known that some of these PDEs could be inhibited (i.e. their catalysing function could be completely or significantly blocked) by certain other chemicals, called PDE inhibitors. A table set out at the back of the primer shows the PDEs, what they acted on and examples of selective inhibitors which were known by June 1993. The relevant data are as follows:

PDE type	Substrate	Inhibitors
PDEI	Hydrolyses cAMP and cGMP	None known
PDEII	Hydrolyses cAMP and cGMP	None known
PDEIII	Hydrolyses cAMP and cGMP	Amrinone (Sterling Winthrop) Milrinone (Sterling Winthrop)
PDEIV	Hydrolyses cAMP	Rolipram (Schering) Ro 20-1724 Denbufylline
PDEV (including PDEV B and PDEV C)	Hydrolyses cGMP	Zaprinast (M&B22948) (although also inhibits PDEI to a lesser extent.)

19. It will be seen that PDE_IV only acts as a catalyst for the destruction of cAMP. It can therefore be said to be a cAMP specific PDE. Similarly PDE_V can be called cGMP specific. The other PDEs are not specific, but can catalyse reactions which destroy both cAMP and cGMP. However, a PDE which catalyses the destruction of both cAMP and cGMP, and therefore is not specific, may be more effective in helping to destroy one type of substrate than the other. For example PDE_III is more effective in destroying cAMP than in destroying cGMP, so it can be called a cAMP selective PDE. Similarly an inhibitor may be more effective in blocking the enzymatic activity of one type of PDE than of another. In such a case it can be said to be a selective inhibitor of the former. So zaprinast, a chemical which figures extensively in this case, is a selective PDE_V inhibitor. In fact the inhibition of an enzyme is a reversible effect which is dependent on the concentration of the inhibitor used. The more inhibitor present, the greater the inhibition. For this reason, one measure of the power of an inhibitor is to find the figure which causes 50% inhibition of the enzyme activity. This is called its IC₅₀ value.
20. It was common general knowledge before the priority date that cGMP and cAMP also regulate a variety of functions in other organs and tissues of the body, not just the relaxation of smooth muscles in the penis.
21. Finally I should mention that the importance of nitric oxide became so well known that in the 18 December 1992 issue of the eminent journal Science it was named "Molecule of the Year". In the editorial of that issue, it was stated that NO "is a major biochemical mediator of penile erection" and in an article in the same issue under the heading "Molecule of the Year", it was said:

"This year [1992], scientists proved definitively that in males, NO translates sexual excitement into potency by causing erections. Key pelvic nerves get a message from the brain and make nitric oxide in response. NO dilates blood vessels throughout the crucial areas of the penis, blood rushes in, and the penis rises to the occasion."

22. One of the Pfizer witnesses, Dr. Challiss, described this as tabloid journalism, and so it is, but he also said it made a nice story. I have no doubt that it reflects the excitement that the elucidation of nitric oxide's role in the body, and in relation to male sexual activity, generated by the end of 1992.

The Patent

23. The patent is for the second medical use of certain compounds. As I have mentioned, one of these is sildenafil citrate, the active ingredient in Viagra. In fact sildenafil citrate and a very large number of the other chemicals covered by the patent are the subject of two earlier Pfizer patent applications, namely EP 0463 756 and EP 0526 004, referred to respectively as Bell I and Bell II. Those applications were published before June 1993 and they are relied on in this case as prior art. In Bell I and II, sildenafil citrate and the other chemicals were proposed for a number of other medical applications, but not the treatment of male erectile impotence. That is why the patent in suit is, in part at least, for second medical use.

(a) The specification – general description of the invention

24. It is particularly important in this case to construe the patent without regard to after acquired knowledge. The first line of the patent states that the "invention relates to the use of a series of pyrazolo[4,3-d]pyrimidin- 7-ones for the preparation of a medicament for the treatment of impotence". This is consistent with the title of the patent which is "Pyrazolopyrimidones for the treatment of impotence." The chemicals which are the subject of the Bell I and II applications are pyrazolo[4,3-d]pyrimidin- 7-ones and they are the subject of some of the claims in the patent, but the patent and its claims cover a much wider field. At page 2 lines 5 to 22 there is a brief discussion of male impotence and some of the prior art methods for treating it. As far as impotence is concerned, the specification says:

"Impotence can be defined literally as a lack of power, in the male, to copulate and may involve an inability to achieve penile erection or ejaculation, or both. More specifically, erectile impotence or dysfunction may be defined as an inability to obtain or sustain an erection adequate for intercourse." (p 2 line 5 to 7)

25. The specification then states that the efficacy of orally administered drugs is low. It refers to the current treatment consisting of i.c. injection of various drugs "either alone or in combination" (page 2 line 15) but it also points out the existence of undesirable side effects, including pain and fibrosis of the penis. Other forms of treatment are touched upon briefly. Again, the specification records that they suffer from side effects. The specification then addresses what it says is the invention in the following terms:

"The compounds of the invention are potent inhibitors of cyclic guanosine 3- 5-monophosphate phosphodiesterases (cGMP PDEs) in contrast to their inhibition of cyclic adenosine 3,5-monophosphate phosphodiesterases (cAMP PDEs). This selective enzyme inhibition leads to elevated cGMP levels which, in turn, provides the basis for the utilities already disclosed for the said compounds in [Bell I] and [Bell II], namely in the treatment of stable, unstable and variant (Prinzmetal) angina, hypertension, pulmonary hypertension, congestive heart failure, atherosclerosis, conditions of reduced blood vessel patency e.g. post- percutaneous transluminal coronary angioplasty (post-PTCA), peripheral vascular disease, stroke, bronchitis, allergic asthma, chronic asthma, allergic rhinitis, glaucoma, and diseases characterised by disorders of gut motility, e.g. irritable bowel syndrome (IBS).

Unexpectedly, it has now been found that these disclosed compounds are useful in the treatment of erectile dys-function. Furthermore the compounds may be administered orally, thereby obviating the disadvantages associated with i.c. administration."

26. A number of points should be noticed. First, at least up to this point in the specification, there is little doubt that the expression "compounds of the invention" and "these disclosed compounds" alike refer back to the pyrazolo[4,3-d]pyrimidin- 7-ones referred to in the opening line which are the subject of the two Bell applications. Secondly, the first paragraph in the last quoted extract asserts that "the compounds of the invention" are both potent inhibitors of cGMP PDEs as compared to their inhibition of cAMP PDEs and that it is this selective inhibition which leads to the elevated cGMP levels said to give rise to the effectiveness of these compounds in the treatment of diseases such as angina, pulmonary hypertension and so on. So, if one looks back at the flow chart set out in Figure 2 above, the compounds of the invention block step 4, namely the ability of the phosphodiesterase enzyme (PDE) to turn cGMP into 5'GMP. The result is that cGMP levels remain high and it is this high level of cGMP which is asserted to be responsible for the medicinal properties of the compounds (i.e. anti angina etc) which are set out in the Bell applications. Thirdly it is said that these compounds 'may be' administered orally. It is not suggested that all of the compounds are suitable for oral administration and it is not Pfizer's position that they all are.
27. The specification then goes on to identify a group of compounds which are said to embody the invention. They are described by reference to a structural formula, referred to as "Formula I", which is set out in Figure 3 below. In it the letters R¹ to R⁴ represent a large number of identified chemical groups. It is not in dispute that the number of chemicals covered by Formula I is very large indeed.

28. The specification then goes on to identify a "preferred group of compounds" within Formula I, a "more preferred group", and a "particularly preferred group", each of which is a smaller subset of the previous group of compounds. Reference is then made to "specially preferred individual compounds of the invention" which "include" 9 named chemicals. One of those is sildenafil citrate.

(b) The specification – experimental data.

29. The specification contains a limited amount of data to explain and support the invention. It explains that preliminary investigations were carried out for the purpose of isolating and identifying the PDEs in human corpus cavernosum and describes the experiments used. It reports that three distinct PDEs were present. They are identified as PDEs _II, _III and _V. The implicit teaching is that PDEs _I and _IV are not present, or not present in detectable amounts. The specification says that the major PDE present is PDE_V. Rather confusingly it states at one point that this PDE is "highly selective" for cGMP (page 4 line 46) while at another it identifies it as "cGMP–specific" (page 4 line 55). There appears to be no doubt amongst the witnesses and the parties that this PDE is PDE_V according to the accepted nomenclature and would be so understood by the skilled man in the art. In accordance with the terminology discussed at paragraph 18 above, it would be best described as cGMP–specific. As far as the second PDE isolated is concerned, the specification states that it destroys both cAMP and cGMP (page 4 line 48) and it identifies it as PDE_II. Although it does not expressly state that this PDE is cAMP selective, it does refer to it as cAMP PDE_II (page 4 line 55). The better view is, therefore, that the reader is being informed that PDE_II destroys both cAMP and cGMP, but is particularly effective against the former. As to the third PDE, the reader is told that it is cAMP selective (page 4 line 49). It is identified as PDE_III. In summary, the reader is being told that there are three PDEs present and that they all destroy cGMP. However the most prevalent PDE is PDE_V which only destroys cGMP.

(c) The specification – what the patented compounds do

30. Having explained that only three PDEs have been detected in human corpus cavernosum, the specification turns to explain how the compounds of the invention work. It says:

"The compounds of the invention have been tested in vitro and found to be potent and selective inhibitors of the cGMP-specific PDEv. For example, one of the especially preferred compounds of the invention has an IC₅₀ = 6. 8 nM v. the PDEv enzyme, but demonstrates only weak inhibitory activity against the PDE_II and PDE_III enzymes with IC₅₀ = >100 µM and 34 µM respectively". (p 4 line 57 to p 5 line 2)

31. The teaching of this passage is that some of the compounds within Formula I have been tested. They strongly and selectively inhibit cGMP-specific PDE_V. This means that they block that PDE much more than the other two PDEs which are also inhibited, but to a much smaller degree. Therefore use of the inhibitor will de-activate much of PDE_V but leave most of PDE_II and _III unaffected. We can see therefore that much of the PDE activity which destroys cGMP, namely that caused by PDE_V, is removed. However some cGMP destroying activity remains because most of the PDE_II and _III activity survives and both of them destroy cGMP to some extent, although less so than they destroy cAMP. To the very small extent that PDE_II and _III are inhibited, their ability to destroy cAMP will also be diminished. Because most of the cGMP destroying activity has been removed, cGMP should rise. This is addressed in the next passage in the specification:

"Thus relaxation of the corpus cavernosum tissue and consequent penile erection is presumably mediated by elevation of cGMP levels in the said tissue, by virtue of the PDE inhibitory profile of the compounds of the invention."

32. This passage is important. The compounds of the invention cause a rise in cGMP levels in the corpus cavernosum, because they prevent some of the cGMP from being destroyed. They also cause the relaxation of the muscle in that tissue which leads to erection of the penis. On the basis of this, the patentee states that the erection is "presumably" caused by the increased cGMP levels. This is, no doubt, a reasonable deduction. But the patentee does not say that he has done anything by way of experimentation to prove this theory is right. It is his explanation of the likely reason why the PDE_V selective inhibitors of Formula I work.
33. The specification then goes on to deal briefly with toxicity, oral availability and efficacy trials. Somewhat surprisingly none of this identifies any particular compound as having been tested. For example it says that patient studies were conducted on one of the especially preferred compounds and showed that it induced penile erection in impotent males. It does not identify which of the nine listed especially preferred Formula I compounds was tested.
34. At page 5 line 15 the patentee states:

"Generally, in man, oral administration of the compounds of the invention is the preferred route, being the most convenient and avoiding the disadvantages associated with i.c. administration."

35. The whole of the specification up to this point appears to have been written by reference to the large number of pyrazolo[4,3-d]pyrimidin-7-ones covered by Formula I and described in the earlier Bell applications. These are the selective enzyme inhibitors referred to on page 2 of the specification. However the specification ends with a paragraph which considerably widens the disclosure and which is of particular significance in this case since it is the basis for the most important claims in the patent. It reads as follows:

"Moreover, the invention includes the use of a cGMP PDE inhibitor, or a pharmaceutically acceptable salt thereof, or a pharmaceutical composition containing either entity, for the manufacture of a medicament for the curative or prophylactic oral treatment of erectile dysfunction in man." (page 5 lines 23 to 25)

36. It should be noticed that, notwithstanding the title to the patent, this passage is not restricted to pyrazolopyrimidones (whether of Formula I or otherwise). Furthermore, unlike earlier passages in the specification, it is not restricted, at least expressly, to selective or specific inhibitors. It appears to be directed at any cGMP PDE inhibitor, whether selective, specific or not. In other words it is directed at any inhibitor which will cause increases in cGMP levels. Thus it appears to cover inhibitors of all the cGMP PDEs listed in the table set out in paragraph 18 above except inhibitors for PDE_IV, because that PDE does not destroy cGMP.

(d) The claims

37. The patent has eleven claims. Attention during the trial was directed to claims 1, 10 and 11, but reference was also made to claims 6, 7 and 9. These claims are as follows:

"Claim 1. The use of a compound of formula (I):



wherein

R¹ is H; C₁-C₃ alkyl; C₁-C₃perfluoroalkyl; or C₃-C₅ cycloalkyl;

R² is H; C₁-C₆ alkyl optionally substituted with C₃-C₆ cycloalkyl; C₁-C₃ perfluoroalkyl;

or C₃-C₆ cycloalkyl;

R³ is C₁-C₆ alkyl optionally substituted with C₃-C₆ cycloalkyl, C₁-C₆ perfluoroalkyl; C₃-C₅ cycloalkyl; C₃-C₆ alkenyl; or C₃-C₆ alkynyl;

R⁴ is C₁-C₄ alkyl optionally substituted with OH, NR⁵R⁶, CN, CONR⁵R⁶ or CO₂R⁷; C₂-C₄ alkenyl optionally substituted with CN, CONR⁵R⁶ or CO₂R⁷; C₂-C₄ alkanoyl optionally substituted with NR⁵R⁶; (hydroxy)C₂-C₄ alkyl optionally substituted with NR⁵R⁶; (C₂-C₃ alkoxy)C₁-C₂ alkyl optionally substituted with OH or NR⁵R⁶; CONR⁵R⁶; CO₂R⁷; halo; NR⁵R⁶; NHSO₂NR⁵R⁶;NHSO₂R⁸; SO₂NR⁹R¹⁰; or phenyl, pyridyl, pyrimidinyl, imidazolyl, oxazolyl, thiazolyl, thienyl or triazolyl any of which is optionally substituted with methyl;

R⁵ and R⁶ are each independently H or C₁-C₄ alkyl, or together with the nitrogen atom to which they are attached form a pyrrolidinyl, piperidino, morpholino, 4-N(R¹¹)-piperazinyl or imidazolyl group wherein said group is

optionally substituted with methyl or OH;

R⁷ is H or C₁-C₄ alkyl;

R⁸ is C₁-C₃ alkyl optionally substituted with

NR⁵R⁶;

R⁹ and R¹⁰ together with the nitrogen atom to which they are attached form a pyrrolidinyl, piperdino, mor-pholino or 4-N(R¹²)-piperazinyl group wherein said group is optionally substituted with C₁-C₄ alkyl, C₁-C₃ alkoxy, NR¹³R¹⁴ or CONR¹³R¹⁴;

R¹¹ is H; C₁-C₃ alkyl optionally substituted with phenyl; (hydroxy)C₂-C₃ alkyl; or C₁-C₄ alkanoyl;

R¹² is H; C₁-C₆ alkyl; (C₁-C₃ alkoxy)C₂-C₆ alkyl; (hydroxy)C₂-C₆ alkyl; (R¹³R¹⁴N)C₂-C₆ alkyl; (R¹³R¹⁴NOC)C₁-C₆ alkyl; CONR¹³R¹⁴; CSNR¹³R¹⁴; or C(NH)NR¹³R¹⁴;

and

R¹³ and R¹⁴ are each independently H; C₁-C₄ alkyl; (C₁-C₃ alkoxy)C₂-C₄ alkyl; or (hydroxy)C₂-C₄ alkyl.

or a pharmaceutically acceptable salt thereof, or a pharmaceutical composition containing either entity, for the manufacture of a medicament for the curative or prophylactic treatment of erectile dysfunction in a male animal, including man.

Claim 6: "The use according to claim 5 wherein the compound of formula (I) is 5-[2-ethoxy-5-(4-methyl-piperazinylsulphonyl)phenyl]-1-methyl-3-n-propyl-1,6-dihydro-7H-pyrazolo[4,3-d]pyrimidin-7-one."

Claim 7: "The use according to claim 5 wherein the compound of formula (I) is 5-(2-ethoxy-5-morpholinoacetylphenyl)-1-methyl-3-n-propyl-1,6-dihydro-7H-pyrazolo[4,3-d]pyrimidin-7-one."

Claim 9: "The use according to any one of claims 1 to 8 wherein the medicament is adapted for oral treatment."

Claim 10: "The use of a cGMP PDE inhibitor, or a pharmaceutically acceptable salt thereof, or a pharmaceutical composition containing either entity, for the manufacture of a medicament for the curative or prophylactic oral treatment of erectile dysfunction in man."

Claim 11: "The use according to claim 10 wherein the inhibitor is a cGMP PDE_v inhibitor."

38. Claim 1 is directed to the use of PDE inhibitors published in the Bell Applications. Although its validity is attacked, Lilly is not so concerned with this claim, or any of the claims dependent on it, because it has no current commercial interest in any inhibitor falling within the scope of Formula I. Claim 6 is directed to the second medical use of sildenafil citrate (Viagra) and Claim 7 is directed to a proposed back-up product to Viagra. Claims 10 and 11 are, potentially, much wider than Claim 1 because they are not limited to a specific series of chemicals. They cover any chemical, whether known now or not, which is a cGMP PDE inhibitor (claim 10) or a cGMP PDE_V inhibitor (claim 11). It is these claims which most concern Lilly.

Construction of the Claims

39. A number of differences exist between the parties as to the proper construction of these claims. However all of the disputes relating to the scope of claims 1, 6, 7 and 9 apply equally to claims 10 and 11 and in relation to the latter there are additional areas of disagreement. For this reason it is more convenient to consider the scope of the latter two claims first.

(a) "The use of … for the curative or prophylactic … treatment of …"

40. Two points arise for consideration here. Pfizer says that these words are only fulfilled by the use of a compound which is both for the purpose of trying to treat the target illness and which also is suitable for treating that illness, that is to say that in relation at least some individuals the treatment works. In support of this it relies on the decision of the Court of Appeal, Bristol-Myers Squibb Co v. Baker Norton Pharmaceuticals [2000] IP & T 908. Furthermore they say that the success of the treatment must be attributable to the use of the identified compounds alone. A successful treatment which involves the use of the identified compounds and other agents in combination, save where the latter are merely colourable, does not fall within the claim. Lilly argue that it is sufficient if the use of the compound is for trying to treat the illness, efficacy is not necessary. It says that if someone uses a product with the intention of treating the illness and is confident it works, then this requirement is met even if, in reality, the product has no relevant pharmacological effect. If the patient's condition improves because he has been told by a doctor that he is going to get better – a placebo effect – then that falls within the claim. For this Pfizer relies on Bristol-Myers Squibb as well. It also disagrees on the issue of combined treatments. If the target compound is successful to treat the illness, it does not matter that such success involves the collaboration of the compound with other agents. In such cases each and all the ingredients in the combined therapy are used for the curative or prophylactic treatment.
41. On the first of these points, I have no doubt that Pfizer's argument is correct. This point was considered by the Court of Appeal in Bristol-Myers Squibb in which Aldous LJ said:

"The words 'for treating cancer' have to be construed in context. The skilled addressee would realise that drugs which were suitable for treatment would not always be successful. However, drugs which had no effect were not suitable. The phrase means 'suitable for trying to treat cancer'. What is suitable is a question of fact, not one of perception. If the drug has a beneficial effect in the treatment of cancer it will be suitable. If not, it will not be." (paragraph 21)

42. A second medical use claim only survives because the compound is effective to achieve a new treatment. If it is not effective, or not discernibly so, it is not suitable for that treatment. If administration of the compound results in some patients getting better, but the same improvement would be achieved by the administration of any placebo, that is not enough. The patient benefit is achieved simply because the patient believes he is being treated, not because of any peculiar feature or efficacy of the patented compound.
43. On the second point, in my view Lilly's construction is to be preferred. When two or more agents are used together to treat a disease then each is "used … for the … treatment of" the disease. Were someone to use a compound within Formula I of the patent in a dose which did not produce full erections in a patient and boosted its performance by the addition of another agent, not within Formula I, I think it would be fanciful to say that the first compound was not being used in the treatment. The point can be made by reference to a real example. It will be recalled that on page 2 of the specification the patentee states that good results in the treatment of MED had been based on the i.c. injection of drugs such as phenoxybenzamine, phentolamine, papaverine and prostaglandin E₁, "either alone or in combination". It would not make sense to say that papaverine, for example, was being used in the treatment of MED when injected alone, but not when it was injected with other agents.

(b) "The use of a cGMP PDE inhibitor" (claim 10) and "a cGMP PDE_V inhibitor" (claim 11)

44. These are the words which have generated the greatest dispute on construction between the parties. Its outcome has a major impact on the experiments which were carried out on the issue of anticipation. Lilly argues that the words should be given their natural unqualified meaning. Therefore any compound which inhibits cGMP PDE falls within claim 10. It does not matter whether it also inhibits cAMP PDEs as well, nor does it matter whether it inhibits cGMP PDEs more effectively than cAMP PDEs. Similarly, in relation to claim 11, Lilly argues that any compound which inhibits cGMP PDE_V is covered, whether or not it also inhibits other PDEs and, in particular, whether or not it inhibits PDE_V more or less than other PDEs. .
45. Pfizer's position on this issue has changed. A dispute arose between the parties as to whether certain experiments should be allowed to be carried out in relation to the issue of anticipation. The matter came on for hearing before Pumfrey J. He ordered Pfizer to serve on Lilly a statement of its case effectively requiring it to explain its construction of claims 10 and 11. This was complied with in June of this year. The Pfizer statement reads as follows:

"Claim 10

In order for its use to fall within claim 10, a substance (or its pharmaceutically acceptable salt, or a composition containing either) must have the following properties:

- It must be a selective cGMP PDE inhibitor, which is to say it must be a potent inhibitor of cGMP PDEs in contrast to its inhibition of cAMP PDEs.

- In addition, the medicament produced using the said substance (or salt or composition) must be for the curative or prophylactic oral treatment of erectile dysfunction in man, which includes the requirement that oral administration should result in an improvement of erectile dysfunction (as compared with the level of dysfunction encountered or to be expected had the substance not been administered). It is not a requirement that the medicament bring about such improvement in all subjects to whom it is administered.

- In addition, the said improvement of erectile dysfunction must be achieved substantially as a result of cGMP PDE inhibition.

Claim 11

In addition to the requirements of claim 10, for the use of a substance to fall within claim 11 it must be a PDE_V inhibitor."

46. Thus, in relation to claim 10, the substance has to be (i) selective, in the sense that it inhibits cGMP PDEs more than cAMP PDEs, (ii) potent in its inhibition, (iii) effective in treating MED and (iv) that effectiveness must have been caused by the inhibition of the cGMP PDE. Claim 11, which is dependent on claim 10, is said to incorporate all these features but is restricted to inhibitors which are selective to PDE_V.
47. This construction gave rise to a number of difficulties. Amongst them was the consequence that, if it was correct, claim 10 only covered selective, potent PDE_V inhibitors and therefore had identical width to claim 11. This was unlikely to be the proper construction and towards the end of the trial Mr Kitchin advanced a somewhat modified view. He said that both claims still required selectivity and potency (and causation) but that claim 10 covered only inhibitors of PDE_I and PDE_V whereas claim 11 covered inhibitors of the latter enzyme only. The way he arrived at this conclusion ran something like this; In 1993 it was thought that cGMP PDE was specific in its effect, that is to say it only destroyed cGMP. It was also thought that PDE_IV was specific to the destruction of cAMP – see the table in paragraph 18 above. It was also known that the other PDEs destroyed both cAMP and cGMP. There was no understanding at that time that PDE_II or PDE_III were in any way selective towards the destruction of cGMP over the destruction of cAMP. However it was said that there was at least one "school of thought" at the time that PDE_I was selective towards cGMP. Mr Kitchin did not rely on this to justify his client's construction of claims 10 and 11. Rather he said that a reader of the patent in suit would realise that the reference to cGMP PDE inhibitor in claim 10 must be a reference back to the discussion of such inhibitors on page 2 of the specification (see the quotation set out in paragraph 25 above) where there is reference to the Bell applications. One would therefore refer to the latter applications. They refer to the presence of PDE_I, PDE_III and PDE_V (under their then current names) and states that the compounds of Formula I are effective anti-angina agents because they inhibit PDE_I and PDE_V. Therefore a reader of the patent in suit would assume that the efficacy of the Formula I compounds in the treatment of MED would similarly be due to the fact that they inhibit PDE_I and PDE_V. From this it would be assumed that similar efficacy could be obtained from any other compound which inhibited PDE_I and PDE_V. Thus the reference in claim 10 to cGMP PDEs must cover any compound (not just Formula I compounds) which can inhibit PDE_I, PDE_V or either of them.
48. This construction at least has the attraction that, unlike others put forward by Pfizer, it appears to give claim 10 a wider scope than claim 11, which is clearly what the draughtsman intended. However, if it were correct, it would have a result which is surprising indeed. As noted already (see paragraph 29 above), the patentee reports that there is no detectable PDE_I in human corpus cavernosum. Therefore claim 10 would, apparently, include medical treatments which involve the inhibition of an enzyme which is not even present in the penis and for that reason presumably would not work. If, of course, such uses are excluded from the claim because they do not work, then one gets back to the situation where claim 10 and claim 11 are of identical width.
49. But in addition to this criticism, it appears to me that Pfizer's approach is wrong in principle. There is nothing in these claims or the specification which suggests that determination of the scope of protection is to be discovered by reference to the Bell applications which are concerned with the treatment of quite different diseases. Furthermore, read in context, Bell I and II tie in with the title of the patent and are identified as the source of the group of chemicals which are the subject of claims 1 to 9. Claims 10 and 11 are based on the much more general wording inserted at the end of the specification (see paragraph 35 above) which has been inserted to widen the disclosure away from the Bell applications. Thus the latter claims are not put forward as based upon or restricted to what is disclosed in those applications.
50. The patentee clearly knew how to refer to specificity, potency and selectivity in relation to PDEs and their inhibitors and did so in the specification where that was his intention. Furthermore there are compelling reasons to construe claims 10 and 11 without the addition of the words 'selective' or 'potent'. The only teaching in the specification which suggests how the patented inhibitors cure or act as prophylactics for MED is that which refers to the inhibition of cGMP PDEs and the consequent rise in cGMP levels. Therefore there is no reason why the patentee should be thought to have avoided seeking protection for second medical uses of any inhibitor which could cause such raised cGMP levels. In other words it makes sense that claim 10 covers any cGMP PDE inhibitor since all such inhibitors will, by definition, increase cGMP levels. The only reason to exclude cGMP PDE inhibitors which also inhibit cAMP PDEs (i.e. non-selective cGMP PDEs) would be if the resultant and simultaneous rise in cAMP levels caused by the use of such compounds would give rise to unacceptable and unavoidable adverse effects. As Mr Kitchin concedes, there is no statement in the specification that raised cAMP levels would be harmful at all, let alone significantly so. This is a point which ties in directly to Pfizer's argument that the claims must cover 'selective' cGMP PDEs. Both parties agreed that there is no magic in the word 'selective'. Mr Kitchin's opening skeleton argument states:

"… persons developing drugs know that they want specific effects and they constantly strive to get them - because non-specific effects tend to give rise to side effects. Hence, by saying a cGMP PDE inhibitor, the patent indicates that that effect rather than any other effect is desired. In fact, of course, the patent also explicitly states that, by contrast to cGMP PDE inhibition, cAMP inhibition is not desired.

The extent to which a compound must inhibit cGMP PDEs in preference to cAMP PDEs to fall within the claim is a question of degree. It is, however, a question of degree which the skilled reader can easily answer; what is desired is a level of selectivity for cGMP PDE inhibition relative to cAMP PDE inhibition so that the former is inhibited enough to give the desired effect, without inhibiting the latter appreciably (thus avoiding non-specific activity).

A secondary point taken by Lilly appears to be that, if the claim does call for selectivity, any degree of preference for cGMP PDE will do, so that a compound with a cGMP PDE:cAMP PDE inhibition ratio of 2:1 would do. Again, this is an approach which ignores the realities. If the ratio were 2:1, then at drug concentrations sufficient to inhibit cGMP PDE enough to get the desired effect, there would also be very substantial, and undesirable, non-specific effects." (emphasis added)

(Note that the last sentence in the first of these three paragraphs is incorrect. There is no such explicit statement.)

51. The whole of this passage emphasises that what would amount to acceptable selectivity is dependent on the existence not just of side effects attributable to the increase of cAMP levels but also the extent to which such side effects are undesirable. If the patentee had intended to make selectivity a feature of his claims, he would have been expected to give the reader some guidance as to the degree of selectivity needed. He did not do so. He did not suggest that there would be side effects if cAMP levels rise, he does not identify any such side effects and, most importantly, he gives the reader no clues as to how rigorously they need to be avoided. Very similar considerations apply to Pfizer's assertion that claims 10 and 11 should be construed as if they contained the word 'potent' before 'inhibitor'.
52. It seems to me that the problem is reinforced by the stance adopted by Pfizer in its closing written submissions where it says:

"Dr Kruse did not agree that cAMP elevation would be dangerous, but thought it might be (XX 4/488-9) and that is good enough for our purposes, since it shows that the skilled man could not read the patent as being indifferent to cAMP elevation (provided that cGMP was elevated). [emphasis as in the original]

53. The fact that the reader of the patent would not be indifferent to cAMP elevation does not mean that he would assume that it must be prohibited at all costs. As Dr Kruse was explaining, elevated cAMP might or might not be undesirable to some extent. The patent includes no indication one way or the other.
54. It follows that I reject Pfizer's suggestion (see paragraph 46 above) that claims 10 and 11 should be construed as if they contained the words 'selective' and 'potent'. For reasons already given, I agree that these claims, and all the others, should be construed so as to cover only such second medical uses as are effective in treating MED. That efficacy involves a degree of potency as a drug. It does not necessarily equate with potency as an inhibitor to cGMP PDEs as compared with cAMP PDEs, which is what Pfizer says the claims mean.
55. This leaves only Pfizer's fourth point, namely that claims 10 and 11 should be construed so as to require the efficacy in treating MED to be caused by the inhibition of the cGMP PDE. This again has impact on the experiments. Pfizer's point is that a use of a cGMP PDE inhibitor does not fall within these claims, even if the use is effective to treat MED, unless it is proved that the mechanism by which the beneficial result was achieved was attributable to that inhibition. Therefore to prove infringement it would be necessary to prove that it was the PDE inhibition which was the cause and not some other activity of the inhibitor. Of course similar proof of causation would need to be provided before any use of a cGMP PDE inhibitor in the prior art could amount to anticipation – hence the impact on the experiments in this case.
56. In my view there is nothing in this point. Nowhere in the claims or the specification does the patentee tie himself down to any particular mode of action of the inhibitor. All that he does (see paragraphs 31 and 32 above) is speculate, no doubt reasonably and probably accurately, that elevation of the cGMP levels caused by the inhibition is the reason for the beneficial effect claimed. If it subsequently were to be found that cGMP PDE inhibitors were effective in curing or reducing MED but by a chemical route other than the elevation of cGMP levels, the claims would have the same width and, in particular, would cover anyone who used a suitable inhibitor to effect the same treatment. Consistent with this, the patent contains no teaching as to what the reader is supposed to do to demonstrate whether or not a particular successful treatment owes its efficacy to elevation of the levels of cGMP. It follows that I do not accept Pfizer's arguments on this point.
57. Finally on construction I should mention one other matter. All the claims are restricted to second medical uses leading to the "curative or prophylactic treatment of erectile dysfunction" in male animals or man. As pointed out at paragraph 24 above, "erectile dysfunction" is defined in the specification as meaning the inability to obtain or sustain an erection. That it is what those words mean in the claims.

Summary on construction:

58. My findings on the construction of the claims are therefore as follows:

(A) Claim 1 covers the use of a Formula I compound (a) either alone or with some other agent, (b) by any means of administration, (c) for the purpose of treating male animals who have the inability to obtain or sustain an erection and (d) where such use is effective, at least in some cases, in treating that disorder.

(B) Claim 10 covers the use of (a) any compound which inhibits cGMP PDE enzymes (whether selective or not), (b) either alone or with some other agent, (c) by oral administration, (d) for the purpose of treating males who have the inability to obtain or sustain an erection and (e) where such use is effective, at least in some cases, in treating that disorder.

(C) Claim 11 is of the same scope as claim 10 save that it is restricted to compounds which inhibit cGMP PDE_V.

Attacks on validity.

59. It is now possible to turn to the attacks on validity. Although anticipation, obviousness, insufficiency and added matter were all relied on by Lilly, the objections under the last two heads were closely tied in to Pfizer's construction of the claims. In the event they are no longer of great importance. Anticipation was advanced on the basis of one document; "Treatment of Vasculogenic Sexual Dysfunction with Pentoxifylline" by Korenman & Viosca - Journal of the American Geriatrics Society (41:363-366) (April 1993). ("Korenman"). Obviousness was pleaded on the basis of a number of prior publications, not including Korenman. However at the trial Lilly concentrated its attention on three documents;

(A) "Nitric oxide as a mediator of relaxation of the corpus cavernosum in respect to nonadrenergic, noncholinergic neurotransmission" by Rajfer et al - New England Journal of Medicine Vol. 362 No. 2 at p. 90 (9 January 1992) ("Rajfer"),

(B) "Phosphodiesterase V_A Inhibitors" by K. J. Murray in Drug News and Perspectives (DN&P) at Vol 6(3) p150-156 (April 1993) ("Murray") and

(C) "The Role of the L-arginine-Nitric Oxide-Cyclic GMP pathway in Relaxation of Corpus Cavernosum Smooth Muscle" a PhD dissertation of Dr Margaret Ann Bush of the University of California ("Bush").

OBVIOUSNESS

60. Obviousness was the major attack advanced by Lilly and I will consider it first. As usual, the parties referred to Windsurfing International Inc. v. Tabur Marine (Great Britain) Ltd [1985] RPC 59. That case sets out a structured approach to the question of obviousness and can simplify analysis. The first step is to identify the inventive concept embodied in the patent in suit. Then one must assume the mantle of the normally skilled but unimaginative addressee in the art at the priority date and to impute to him what was, at the date, common general knowledge in the art in question. Third, it is necessary to identify what if any differences exist between the matter relied on as being "known or used" and the alleged invention. Finally, one must ask whether, viewed without knowledge of the alleged invention those differences constitute steps which would have been obvious to the skilled man or whether they require any degree of invention. Windsurfer is a particular help where the inventive concept is buried in obscure language in the claims. Here that is not so. Mr Kitchin set out the relevant concepts as follows: claim 1 is to the use of the compounds of Formula I for the manufacture of medicaments for the treatment of MED. Note that there is no limitation to any particular form of administration. Claim 10 is to the use of a cGMP PDE inhibitor for the manufacture of a medicament for oral treatment of MED. So claim 1 covers any mode of administration, but is restricted to Formula I compounds only whereas claim 10 covers use of any cGMP PDE inhibitor but only if administered orally. Both claims cover sildenafil citrate but, because of its much greater width, the more important claim is the latter (and claim 11).
61. A considerable volume of evidence was advanced in support of and to counter the case on obviousness. Central to this exercise was the evidence given by a number of impressive experts including a Nobel Laureate, Professor Ignarro, who was called on behalf of Pfizer. In the light of the heavy reliance placed by the parties on their respective lead experts' evidence, I think it might be useful to identify and restate the task facing the court.
62. The question of obviousness has to be assessed through the eyes of the skilled but non-inventive man in the art. This is not a real person. He is a legal creation. He is supposed to offer an objective test of whether a particular development can be protected by a patent. He is deemed to have looked at and read publicly available documents and to know of public uses in the prior art. He understands all languages and dialects. He never misses the obvious nor stumbles on the inventive. He has no private idiosyncratic preferences or dislikes. He never thinks laterally. He differs from all real people in one or more of these characteristics. A real worker in the field may never look at a piece of prior art – for example he may never look at the contents of a particular public library – or he may be put off because it is in a language he does not know. But the notional addressee is taken to have done so. This is a reflection of part of the policy underlying the law of obviousness. Anything which is obvious over what is available to the public cannot subsequently be the subject of valid patent protection even if, in practice, few would have bothered looking through the prior art or would have found the particular items relied on. Patents are not granted for the discovery and wider dissemination of public material and what is obvious over it, but only for making new inventions. A worker who finds, is given or stumbles upon any piece of public prior art must realise that that art and anything obvious over it cannot be monopolised by him and he is reassured that it cannot be monopolised by anyone else.
63. Of particular importance in this case, in view of the way that the issue has been developed by the parties, is the difference between the plodding unerring perceptiveness of all things obvious to the notional skilled man and the personal characteristics of real workers in the field. As noted above, the notional skilled man never misses the obvious nor sees the inventive. In this respect he is quite unlike most real people. The difference has a direct impact on the assessment of the evidence put before the court. If a genius in a field misses a particular development over a piece of prior art, it could be because he missed the obvious, as clever people sometimes do, or because it was inventive. Similarly credible evidence from him that he saw or would have seen the development may be attributable to the fact that it is obvious or that it was inventive and he is clever enough to have seen it. So evidence from him does not prove that the development is obvious or not. It may be valuable in that it will help the court to understand the technology and how it could or might lead to the development. Similarly evidence from an uninspiring worker in the field that he did think of a particular development does not prove obviousness either. He may just have had a rare moment of perceptiveness. This difference between the legal creation and the real worker in the field is particularly marked where there is more than one route to a desired goal. The hypothetical worker will see them all. A particular real individual at the time might not. Furthermore a real worker in the field might, as a result of personal training, experience or taste, favour one route more than another. Furthermore evidence from people in the art as to what they would or would not have done or thought if a particular piece of prior art had, contrary to the fact, been drawn to their attention at the priority date is, necessarily, more suspect. Caution must also be exercised where the evidence is being given by a worker who was not in the relevant field at the priority date but has tried to imagine what his reaction would have been had he been so.
64. This does not mean that evidence from those in the art at the relevant time is irrelevant. It is not. As I have said, it may help the court to assess the possible lines of analysis and deductions that the notional addressee might follow. Furthermore, sometimes it may be very persuasive. If it can be shown that a number of ordinary workers in the relevant field at the relevant time who were looking for the same goal and had the same prior art, missed what has been patented then that may be telling evidence of non-obviousness. This is particularly the case where the commercial benefits of the development would have been apparent and a long time had passed between the publication of the prior art and the priority date of the patent. Hence the impact and interrelationship between the familiar concepts of long felt want and commercial success. Likewise evidence that ordinary men in the art and working from the same prior art at the relevant time independently came to the same development may be some evidence that the notional skilled man would have done likewise. However the evidence is rarely that simple. In most substantial patent cases the technology at issue is sophisticated and the witnesses called are experts in their fields. In most cases, of which this is a good example, they are either renowned academics or researchers who have been immersed in the Research and Development departments of major companies. In either case they come to the issues not only with a profound understanding of the technology but also frequently with knowledge of additional private and relevant information which is not deemed to be known to the notional addressee. For example a research worker in the field will almost always have knowledge of highly confidential prior work done in his department which cannot but affect his attitude to the prior art. This is all material of which the notional man skilled in the art will be ignorant.
65. The actions of the notional skilled man also reflect on a topic sometimes called by British patent lawyers "mosaicing". To what extent is it possible to patch together the teaching or disclosures from different sources? On this topic I was reminded of what is said in Terrell, 15^th Edition, starting at paragraph 7.13:

"Each document must be interpreted on its own in order to determine the information it contains. It is not legitimate to piece together a number of prior documents in order to produce an anticipation of the invention. … However, a series of papers which form a series of disclosures and refer to each other, so that 'anyone reading one is referred by cross-reference to the others', do not form an impermissible mosaic and can be relied upon together as an attack on novelty."

66. This passage is directed particularly at the issue of mosaicing when applied to the law of novelty. The same approach applies to obviousness. There may well be invention in patching together disclosures from unrelated sources (see Von Heyden v. Neustadt (1880) 50 LJCh 126). But, at least in relation to obviousness, the second part of this statement does not represent a rigid but limited exception. When any piece of prior art is considered for the purposes of an obviousness attack, the question asked is "what would the skilled addressee think and do on the basis of this disclosure?" He will consider the disclosure in the light of the common general knowledge and it may be that in some cases he will also think it obvious to supplement the disclosure by consulting other readily accessible publicly available information. This will be particularly likely where the pleaded prior art encourages him to do so because it expressly cross-refers to other material. However I do not think it is limited to cases where there is an express cross-reference. For example if a piece of prior art directs the skilled worker to use a member of a class of ingredients for a particular purpose and it would be obvious to him where and how to find details of members of that class, then he will do so and that act of pulling in other information is itself an obvious consequence of the disclosure in the prior art.
67. Finally on the issue of the characteristics and behaviour of the skilled addressee, mention should be made of the skilled team. For many years now it has been well accepted law that in some cases the addressee for the purpose of testing obviousness is considered to be a team made up of notional skilled but uninventive members from different disciplines. Although both parties accepted that this was a case where the skilled addressee should be considered to consist of a team, at one stage it appeared that there was a dispute between them as to the nature of the expertise of the members of the team. Both sides accept that the relevant fields of expertise would encompass pharmacology, medicinal chemistry and urology. However Pfizer suggest that, at least at first, the notional team would only have a very general background knowledge of PDEs rather than the knowledge of a skilled but unimaginative worker in the PDE field. On the other hand Lilly says that such expertise should be attributed to the notional team from the outset. In the end there was no substantial difference between the parties on this topic. As Mr Kitchin put it in his skeleton:

"…one should consider [the notional team's] reaction to the prior art item by item, but bearing in mind that if the art specifically flags a technology in which they would regard themselves as inadequately skilled, they would consider getting help from someone else."

68. I have come to the conclusion that the notional skilled addressee or team in this case would have relevant but unimaginative expertise and knowledge in the fields of pharmacology, chemistry and urology and that would include knowledge of most treatments for MED which were publicly available at the priority date in June 1993 together with non-inventive expertise in relation to PDEs.

(A) Rajfer

69. This document was described as creating a furore and being very exciting. Professor Ignarro said that it had played a role in making Science magazine declare nitric oxide the Molecule of the Year (see paragraph 21 above). It is the product of work carried out by Professor Ignarro's team and it lists Professor Ignarro, Dr Bush (of the Bush Thesis) and others in addition to Dr. Rajfer as authors. It is not in dispute that the skilled reader at the time would have thought of this as reporting an important and high quality piece of research.
70. The brief extract at the beginning of the paper reports:

"The relaxation [of human corpus cavernosum tissue] caused by either electrical stimulation or nitric oxide was enhanced by a selective inhibitor of cyclic guanosine monophosphate (GMP) phosphodiesterase (M&B 22,948)" [i.e. by a selective cGMP PDE inhibitor]

71. Under the heading 'conclusions' in the abstract, the authors say that defects in the NANC pathway may cause some forms of impotence. The main text of the paper records that failure of penile erection could be due to impaired relaxation of the smooth muscle of the corpus cavernosum (p. 90 left column) and states that the purpose of the study it reports was to ascertain whether nitric oxide plays a part in relaxation of the corpus cavernosum in humans and therefore in penile erection (p. 90 right column). It then reports a series of experiments. It is not necessary to go into the detail of those experiments but they can be split into two broad groups. In the first, the effect of modification of the nitric oxide generating part of the NANC pathway was examined. For example the generation of nitric oxide was impaired by swamping the tissue samples with an analogue of L-arginine which is not turned into NO. This had the effect of preventing L-arginine from being turned into NO (see step 1 in the flow diagram at paragraph 16 above). The second was concerned with looking at the chemical processes by which the cGMP produced by NO was removed from the tissue. To do this it used a cGMP PDE inhibitor identified as M&B 22,948 – this is the pharmaceutically active molecule known as zaprinast produced by May & Baker (see the table at paragraph 18 above). The authors report that they took strips of fresh human corpus cavernosum tissue and subjected them either to electrical field stimulation (i.e. to emulate the effect of the nerves when a man is sexually aroused) or to treatment with a chemical called SNAP which liberates NO. In both cases the NO released made the corpus cavernosum relax. The authors then applied the selective cGMP PDE inhibitor, zaprinast, to the tissue samples to see what effect it had on the NO-generated relaxation. It records that the inhibitor augmented or enhanced the relaxant responses. For example under the rubric "Preliminary Observations in Men without Impotence" Rajfer says:

"Corporal tissue from two men without impotence and tissue from impotent patients responded similarly with regard to the … enhancing effects of M&B 22,948 on electrically elicited relaxation. … M&B 22,948 … caused increases of 72 to 86 percent and 84 to 96 percent, respectively, in the relaxation response of corporal strips from 2 normal patients and 11 impotent patients."

72. In the discussion section of the paper, Rajfer says:

"… the addition of nitric oxide … caused similar rapid relaxant responses that were … enhanced by M&B 22,948 … M&B 22,948 is a selective inhibitor of cyclic GMP but not cyclic AMP phosphodiesterases…" [i.e. it is a selective inhibitor of cGMP PDE but not cAMP PDE]

and ends by concluding that the NANC pathway

"…may be involved physiologically in mediating penile erection. It is conceivable that impairment of this pathway could account for the impairment in relaxation elicited by electrical-field stimulation that has been described in certain impotent men. Smooth-muscle relaxation is the mechanism by which papaverine and prostaglandin E₁, when injected intracavernosally, cause tumescence in impotent men. … interference with the L-arginine-nitric oxide pathway could be one cause of impotence that is treatable by the administration of direct-acting vasodilators."

73. There is one other passage in Rajfer which should be referred to. It is on the second page of the paper in the detailed description of the experimental procedure being used. It reads as follows:

"When [zaprinast], an experimental drug prepared by May and Baker (Dagenham, United Kingdom) was tested, it was added to the strips at a concentration of either 1 or 3 (mol per liter (whichever caused a relaxation of 10 to 15 percent, sufficient to ensure that enough [zaprinast] had been added)."

74. Mr Thorley argues that a skilled team would have appreciated the significance of the teaching that use of a selective cGMP PDE inhibitor could enhance or augment the erectile response. It would have taken that as an encouragement to try such inhibitors as candidates for medical treatment of impotence, including impotence consisting of an inability to sustain a complete erection. Once it had reached that conclusion it would have been obvious to try any known cGMP PDE selective inhibitor – including any of those in the Bell applications (so that claim 1 and its dependents are obvious). Furthermore it would have been obvious to try to use any cGMP PDE orally (so that claims 10 and 11 are obvious).
75. He also argues that the proof of the pudding is in the eating. On seeing the Rajfer paper a number of individuals in the art thought of using these inhibitors to treat impotence. Thus it is known from Pfizer internal documents that a Dr Ringrose of Pfizer read Rajfer and annotated it in manuscript with the words "Should we not try out [sildenafil citrate] in impotence? Have we seen any beneficial S[ide]/e[effect]s." Similarly we know from Dr. Murray's paper which is the second main piece of prior art that he, a senior biologist at SmithKline Beecham, had thought of it, as had a Dr. Silver of Sterling Winthrop, Dr. Gristwood of Almirall and Dr. Hyafil of Glaxo. To these should be added Dr. Bush who made the same suggestion in her thesis. This is an impressive list. However only Drs Gristwood and Hyafil gave oral evidence. Dr Bush provided a witness statement but she was not cross-examined. It is likely that all of these individuals were highly qualified researchers. It may be that some of them were inventive people. It is apparent that all, with the possible exception of Professor Ignarro and Dr. Bush, were involved in the research departments of major pharmaceutical companies. It may be that their reaction to the Rajfer paper was conditioned by the confidential information in this field which they may have had access to inside their respective companies. As far as Dr Hyafil is concerned, his deduction was not made on the basis of the Rajfer paper itself but from hearing it described by Professor Ignarro at a public lecture. Having heard the evidence of the Professor, it is likely that the lecture gave no further relevant information than is to be found in the paper, but one cannot now be sure. In view of these uncertainties, I think it is not possible to say that these actions by other parties prove the case of obviousness, they are merely not inconsistent with it.
76. Mr Kitchin's response to the case based on Rajfer involves a number of submissions. Most of them apply equally to all the claims in issue. However one is directed to demonstrating the validity of claim 1 and its dependents while another is directed to claims 10 and 11. It is convenient to consider them one by one.

(a) Alternative routes to a treatment

77. Mr Kitchin says that there was no reason to start with the L-arginine-NO-cGMP pathway. There were many alternative routes which could have been of interest. For example other forms of i.c. injection might prove interesting, because it would amount to a refinement of an existing successful treatment. Or it might be that someone would be more interested in trying to develop a topically applied treatment.
78. Even were Mr Kitchin's first point good on the facts, it is wrong in law. The fact that there are alternative routes is no answer to a case of obviousness based on a particular piece of prior art. On the contrary, the notional skilled addressee is expected to have read the pleaded prior art carefully and to bring to it his interest in the field. If he does that and finds the patented step was an obvious one to make, it is no answer to say that if he had started with other prior art other solutions would have come to mind. Furthermore this argument does not succeed on the facts. The reaction to Rajfer was explained by Pfizer's witness, Professor Ignarro:

"I recall that shortly before publication of the [Rajfer] article the editor of the New England Journal of Medicine warned me of the likely furore and said that impotent men would be asking where could they get nitric oxide. The paper was front page news in the New York Times for 9 January 1992. For two days we were bombarded, and I got interviewed on TV, on Good Morning America or some equivalent of the time. I declined an interview with Hustler magazine."

79. In my view that popular response would have been mirrored by a skilled man in the field in the sense that he would realise that now a quite different cure for some forms of impotence might be available. What Rajfer did was elucidate the whole pathway including the final step involving the use of cGMP PDE. As Professor Ignarro commented during his cross examination, in the 1980s, the pathway responsible for causing smooth muscle relaxation was known as the NANC pathway. He said that these were:

"… very interestingly named for what they are not, which always impressed me, not adrenergic and not cholinergic, nobody knew what they were so they were called nonadrenergic, noncholinergic nerves." (Day 3 Transcript page 352)

What Rajfer did was elucidate the whole pathway. As Mr. Pryor said, he nailed the pathway and completed the jigsaw. (Day 4 Transcript page 590)

80. Once a worker understands what is going on in a process it is possible to make logical decisions on how to tackle defects. He is no longer restricted to the pharmaceutical equivalent of banging on the side of the set when the television does not work. Rajfer not only elucidated the full pathway but also told the reader that the vital smooth muscle relaxing agent, cGMP, could be increased either by increasing NO production or by preventing or reducing its destruction by cGMP PDE. There is no doubt that the treatments available in the early 1990's were either of low efficacy or unpleasant. The need for an alternative route was apparent. It follows that if developing Rajfer was obvious, it does not cease to be so because of the existence of other possible routes.

(b) Only the front end of the NANC pathway would have been of interest

81. Mr Kitchin also argues that if one did seek to manipulate the L-arginine-NO-cGMP pathway in impotent males, the "only logical thing to do", to use his words, would be to administer a treatment which would increase NO production – i.e. boost the front end of the pathway. There was no reason to think that there was a functioning NO-drive in impotent males so interfering with cGMP PDE would be pointless – there might be no cGMP to enhance.
82. In my view this point is also bad both in law and on the facts. I am not persuaded that the notional skilled worker in the field would only consider the front end of the NANC pathway. What was apparent in the evidence filed in this case was that some workers would, for understandable reasons, prefer to look at the downstream end of the pathway (i.e. the PDE end) and others favoured the upstream end (i.e. the NO production end). Professor Ignarro was the most determined of the latter group. I will consider his evidence with some care when I come to consider the Bush thesis. However the fact that there were alternative ways of looking at the problem and that some real life workers preferred to consider one course rather than the other does not answer a case of obviousness. Tied in with this part of Mr Kitchin's argument is the assertion that there was no reason to think that there was functioning NO production in impotent males. If there was no such production, then cGMP would not be produced. If that was what was happening, then there was no cGMP for the cGMP PDE to destroy and inhibiting the latter PDE would be ineffective. This was put by way of an analogy by Dr. Ellis, one of Pfizer's witnesses. He likened the process of production and destruction of cGMP to filling and emptying a bath. The production of NO was equivalent to opening the tap. The PDE which destroys the cGMP is like the plughole which lets the water out of the bath. A cGMP PDE inhibitor therefore acts to put the plug back in. As he put it "there is no use putting in the bath plug unless you turn on the taps as well". Therefore if the breakdown of the NANC pathway was at the NO-production end, using a PDE inhibitor would be pointless. The important word here is 'if'. As all the witnesses appeared to agree, it was not known where the breakdown in the NANC pathway (assuming that to be the cause of the impotence) was in impotent males. Furthermore it appears to have been accepted that there were likely to be different defects in different impotent males. It seems to me that the notional skilled worker in the field at the priority date would have thought it likely that in some impotent males it would be likely that the breakdown was caused by production of cGMP in insufficient quantities, for example because NO was produced but in less than necessary quantities, or for too short a period or by defects in the control of cGMP breakdown. It should be remembered that MED includes those cases where the patient achieves a full erection but is unable to sustain it or only achieves partial erection. In all these cases, using an inhibitor to stop or slow down cGMP breakdown might well result in higher levels of cGMP and therefore more successful erections. It was worth a try.
83. To use Dr Challis' analogy, it was quite possible that in some cases there was no water being poured into the bath because the tap was turned off. But in other cases it was just as likely that the tap was only partly on so that insufficient water entered the bath or the size of the waste outlet was too big, or both of these. It would only be where the tap was completely closed that blocking the outlet would be expected to have no effect. When that analogy is translated to MED, there was no reason to believe, and no witness asserted that it was believed, that all cases of MED attributable to failure of the NANC pathway were cases in which there was total failure of NO production.
84. Mr Kitchin develops his argument further. He says that even if the worker in the art thought of trying to use a cGMP PDE inhibitor, and assuming he had got through the sort of successful in vitro experiments used by Rafjer, he would first consider injecting the chosen agent into an animal subject. For example he would try injecting it into anaesthetised dogs or monkeys. Such treatment works in relation to the injection of nitrovasodilators because they cause an increase in NO production and a consequent increase in cGMP levels. However the base level of NO production in the penis is very low in anaesthetised dogs, so there is likely to be little or no cGMP for the inhibitor to enhance. No erection would be caused and the course of research would be brought to an abrupt end. This is not an unreal suggestion because it is what Pfizer did at first and it produced no discernible effects in the subject animals.
85. In my view this is far too bleak a picture. One of the things that Rafjer teaches is that, unlike nitrovasodilators, which generate cGMP, the cGMP PDE inhibitors only help to potentiate the cGMP which is already there. If there is no basal level of cGMP, then the inhibitors will have nothing to enhance. That is why in many of the experiments reported in the paper, the strips of corpus cavernosum were first subject to electrical field stimulation or SNAP before zaprinast was applied to them. This was to ensure that there was NO present. That message would have been understood by a skilled worker in the field. In fact the passage in Rajfer set out at paragraph 73 above discloses that some strips of corpus cavernosum relaxed 10% to 15% when treated with zaprinast even when no NO source was supplied, presumably because there was base level of NO production in the untreated tissue. This point might well have been missed by the skilled addressee. As we shall see, it was not missed by Dr. Murray. Nevertheless the message that zaprinast or other similar inhibitors would only work if sufficient NO was already present would have been understood by the man skilled in the art. The man skilled in the art would have realised this before he started injecting anaesthetised animals because those subjects are unlikely to be producing NO because they are not sexually stimulated. Furthermore, even if he made the mistake of injecting anaesthetised animals, he would not be put off by failure to see penile erections. He would immediately realise that the likely cause of the negative result was the absence of anything for the inhibitor to enhance. Evidence touching on this subject was given by Mr. Pryor. He said:

"MR. KITCHIN: Could I please take you on to page 181, where you will see a summary. Could I draw your attention, please, to the right-hand column. The author writes: "The development of effective pharmacological means of inducing erections has revolutionized the treatment of erectile dysfunction, but has in no way produced a panacea." By inducing erections, am I correct in understanding that the therapies used produced an erection on administration?

A. Yes, within the limitation that that was not always the case. By that stage [i.e. 1993], it was accepted that if you wanted to maximize the response, you needed some form of sexual stimulation as well." (Day 4, Transcript page 563)

86. How, then, does this marry up with what happened in Pfizer? In early 1992 it injected sildenafil citrate into anaesthetised monkeys. Mr. Kitchin suggests that if a major pharmaceutical company with a considerable research department failed to see the necessity for some form of sexual stimulation in the animal models, so surely would the skilled and non-inventive man in the art. Even were this the totality of the Pfizer story, it would not disprove Lilly's case. Sometimes large companies make mistakes. But the full story is not as simple as it may seem. What happened is that Pfizer had decided in 1990 to look at sildenafil citrate for potential use as a drug. Initially it thought that the compound might have application in the treatment of, amongst others, intermittent claudication, transient ischaemic attacks, hypertension (all circulatory problems) and osteoporosis. According to Dr. Ellis, by August 1991, that is to say well before the publication of Rajfer, sildenafil citrate had been informally scheduled for evaluation in a model of erectile function. At about that time, it was decided to test the compound in monkeys. The nature of the experiments to be undertaken were explained by Dr Ellis:

"the Urogenitals group were working on impotence and were investigating novel (-blockers, injected intracavernosally, for the treatment of MED. … Gorm Wagner was a sex therapist who had developed a model in the monkey to assess compounds as potential treatments for MED. The model involved the use of an anaesthetised monkey. A ligature was placed around the penis and the compound under test injected intra cavernosally. After a few minutes to allow absorption of the compound the ligature was released and any erections recorded. Compounds acting directly to induce erections, such as (-blockers and prostaglandins, are effective in this model inducing full and maintained erections."

87. Therefore the experimental model used was designed to find direct inducers of erections. In those circumstances it was of no consequence that the animal was anaesthetised since it was not required to supply any sexual input. The direct acting compounds would mimic that input. Sildenafil citrate was added to the (-blockers being tested. In other words it was being tested to see if it also was a direct acting compound. One week before these experiments were due to be carried out Dr. Ringrose saw the Rajfer paper for the first time and circulated it with the manuscript note referred to in paragraph 73 above. There is no suggestion that anyone on the team had any knowledge of the details of the NANC pathway before Rajfer was forwarded by Dr Ringrose nor was it suggested that anyone on the team considered the contents of that paper either in detail or at all.
88. The following week the experiments were conducted. Dr Ellis explains that:

"[sildenafil citrate] did not work in this model, inducing only a very transient partial erection …. This indicated that the basal nitric oxide drive was insufficient for sildenafil to be effective in this model. We were disappointed at this result and did not have the conviction to continue exploring the utility of [sildenafil citrate] in MED in the absence of other supportive data. Indeed, I do not recall seeing any formal report of this study."

89. The reason for the failure of this model is then explained by Dr Ellis:

"Subsequently we discovered that an adequate supply of nitric oxide, such as that released during sexual arousal, is essential for a cGMP PDE inhibitor to work. Thus the mechanism of action of cGMP PDE inhibitors is fundamentally different from other (directly acting) vasodilators. The fact, now appreciated, that the production of cGMP is not increased, but the breakdown of cGMP is countered, means that the drug does not in itself cause an erection (like the injected vasodilators) but only enhances the natural erectile response to sexual arousal. Thus, it simply enhances the natural response to arousal. It is essential that this level of nitric oxide is achieved in patients with MED for cGMP PDE inhibitors to be effective. This makes it clear in hindsight, (which it was not at the time) why [sildenafil citrate] did not prove effective in the monkey model."

90. The failure of the experiment is understandable. What is clear from the evidence is that the Pfizer workers involved did not know, to use Dr Ellis' words, that the mechanism of action of cGMP PDE inhibitors is fundamentally different from other direct acting vasodilators. They had not appreciated that the production of cGMP is not increased by an inhibitor but that its breakdown is countered. However these are matters which a skilled man in the art would have understood from Rajfer (and from Murray or Bush). It should be remembered that even the abstract at the beginning of the Rajfer article points out the different mode of action because it states that

"The relaxation caused by either electrical stimulation or nitric oxide was enhanced by a selective inhibitor of cylcic guanosine monophosphate (GMP) phosphodiesterase." (emphasis added)

91. That message is reinforced by the rest of the document. It explains that the inhibitor does not produce the muscle relaxing factor (cGMP) but enhances or amplifies its action if it is already there. Dr Ellis is correct to say that the different mode of action of cGMP PDE inhibitors to NO donors makes it clear why the experiment did not work. However once Rajfer (or Murray or Bush) was published and had been read carefully, that difference was no longer a matter of hindsight.

(c) Only a combined treatment would be thought of

92. Mr Kitchin's third argument goes as follows: if one thought of using a cGMP PDE inhibitor one would only think of doing so in combination with an NO donor. This is because, again, the notional skilled worker would assume that more NO needed to be produced in the corpus cavernosum in all impotent men. So the tap had to be turned on as well as the plug put in the bath. I did not understand this to amount to a concession on behalf of Pfizer that a combined treatment would be obvious.
93. In my view, this argument also fails on the facts and the law. First, it appears to me that if a man in the art thought of harnessing the knowledge provided by the Rajfer paper to the development of a treatment for impotence, he would have considered both ends of the NANC pathway alone and together. The idea of using two or more agents together is neither new nor surprising. As noted above, the patent itself refers to the fact that existing treatments for MED included administration of drugs in combination. This was touched upon by Dr. Ellis under cross-examination:

"A. Sure. The phentolamine story started in the late 80s when there was a very small trial published in the American literature suggesting that phentolamine, taken orally, might be an effective treatment. I think they studied 20 patients at most. At about that time phentolamine was being widely used as an injectable although, interestingly, as an injectable it is not very effective when it is used on its own. It does not produce a very good erection and historically it has always been combined with something else. Initially it was papaverine, subsequently prostaglandin E1 and laterally VIP." (Day 6 Transcript page 769).

94. No-one offered a reason why a combined treatment would not be acceptable. Nor was any reason put forward as to why a man skilled in the art would have failed to think of such a treatment after reading Rajfer. On the contrary, it appears to me that Professor Ignarro came close to saying that a combined treatment, at least applied intracavernosally, was something which would have been expected to have occurred to him

"Thus I never thought at that time that a selective cGMP PDE inhibitor would have any use on its own in the treatment of impotence. I would have thought for this reason that for it to be of any use it would have to be combined with an nitric oxide donor, but that to administer such a combination orally (or for that matter either component on their own) and thereby systemically, with the attendant loss in blood pressure would risk killing the patient. I was surprised to hear that erectile dysfunction could be effectively treated by the oral administration of a cGMP PDE inhibitor."

95. The Professor's references to 'killing the patient' which was tied in to his worries about oral administration of PDE inhibitors will be considered below. For the purpose of dealing with this part of Mr Kitchin's submissions, it is sufficient to note that there appears to be no material which undermines the view that Rajfer would suggest, at the least, a combined treatment of a NO donor and a PDE inhibitor applied intracavernosally. However, as explained above, the claims cover the use of PDE inhibitors in combined treatments, so this is not an answer to the argument of obviousness. But more than this, I have already explained that it would have been apparent to the skilled worker that some impotent males were likely to have functioning or partially functioning NO production in their corpora cavernosa and that the administration of a cGMP PDE inhibitor alone might well work.

(d) No reason to find the Bell applications

96. Mr Kitchin's fourth argument is directed at claim 1 and the claims dependent on it. Although these claims are broad in the sense that they are directed to any mode of treatment, including oral and intracavernosal administration of the drug, they are restricted to the very large class of chemicals encompassed by Formula I. Mr Kitchin says that even were it obvious to think of trying PDE inhibitors for the treatment of MED, there was no reason for a skilled worker to stumble on the Bell applications – which include all or substantially all of the Formula I PDE inhibitors.
97. I reject this argument also. It should be borne in mind that Mr Kitchin conceded, as he had to, that none of the claims in the patent were said to cover a selection invention. In particular, he conceded that he could not say that the act of selecting the vast number of chemicals covered by Formula I out of the even larger number of possible PDE inhibitors was inventive. Not only is there nothing in the patent to support an invention based on selection, but any such argument would be inconsistent with claim 10 which covers all cGMP PDE inhibitors, as long as they are capable of being taken orally. If this is correct, and it is obvious in the light of Rajfer to try to use cGMP PDE inhibitors to treat impotence, it is obvious to try any such inhibitor and there is no invention in listing some but not all of the possible inhibitors. In these circumstances it is obvious to try all the inhibitors including, but not limited to, those disclosed in the Bell applications. It follows that it is not necessary to show that a man in the art would have searched for and found the Bell applications. Nevertheless, if that had been a requirement, it would have been met here. It is not in dispute that a literature search at the priority date for PDE inhibitors would have turned up at least Bell I. It would have been obvious to a skilled man to conduct such a search and to limit it, in the initial stages, to a patent search so as to see what other pharmaceutical companies were doing since it was apparent at the priority date that a considerable number of such companies were engaged in PDE research. Again, a search so limited would have turned up Bell I.

(e) Oral administration would be shunned

98. Mr Kitchin argues that there were strong reasons to avoid oral administration. This was directed to the validity of claims 10 and 11. He says that it would have been apparent to a skilled worker that the tissue concentrations in the penis which he would have assumed were necessary to get an effect were likely to be so high that oral doses would be very large. Furthermore he says that because nitric oxide, cAMP and cGMP are to be found all over the body, to give a PDE inhibitor systemically in large doses would be highly likely to lead to most undesirable, widespread and dangerous effects. Such effects would include falling blood pressure. Hence Professor Ignarro's fear of "killing patients". In addition to this, Mr Kitchin says that there was no established practice of giving oral treatments for impotence; quite the opposite.
99. In my view none of these arguments bears close scrutiny. The starting point in considering this issue is an awareness of what those in the art thought of oral treatment of MED at the priority date. In my view this was an area in which there was close to concensus between the witnesses. There is no doubt that a high priority in the search for any new treatment of MED was that it should be administered orally. This was considered to be the ideal form of treatment by Dr. Gristwood. Dr. Challis, said that oral delivery was clearly the preferred route of administration and, under cross-examination, he said:

"…if you can achieve oral administration I am sure that there are a myriad of reasons why you would pursue it" (Day 5 Transcript page 659 –660)

100. Some of those reasons were given by Dr. Kruse in a passage in his report which was not seriously challenged:

"The ultimate aim of most drug research program is to produce a drug which can be taken orally. … Oral administration is the preferred and most widely used route of administration for the simple reason that it best ensures patient compliance and can be used at home, as opposed to in a hospital setting. Oral drugs can easily be self-administered by a patient. Swallowing a pill is a relatively non-invasive and non-threatening event for most patients and it is a fairly easy task to ensure one is taking the correct quantity of the drug, i.e. it is easier to count pills than say measure out a quantity of solution to take intravenously. Of course there are some circumstances where other routes of administration (e.g. intravenous or intramuscular) might be preferred, for example for the administration of a very potent, short acting drug following a stroke or an anaphylactic shock. Non-oral methods of administration might also be preferred for certain organs, for example administration directly into the eye for the treatment of glaucoma. However, in the vast majority of cases of drugs being administered to treat non-life threatening conditions the oral route of administration is the best and the most widely used. In a disease such as erectile dysfunction, where male self-image, and perceptions of performance are important, oral activity is almost certainly a requirement for a commercially successful drug. It is difficult to imagine an injectable or other mode of administration competing effectively with the discretion and simplicity of an orally acting therapeutic."

101. Similar evidence was given by Mr Pryor, a urologist called by Lilly who was a most impressive witness. He said that by early 1993 the vast majority of clinicians had recognised the disadvantage of penile injection therapy and the desirability of having an oral preparation and that he would have advised any company involved in finding a new treatment for MED that it should make it a priority to develop an orally active drug. He said:

"The oral route of administration is generally the most convenient and most acceptable to a patient for any drug. It enables the patient to manage easily and in the safest way possible the treatment of his particular disorder. It is suitable for acute, chronic or prophylactic treatment. With respect to the treatment of erectile dysfunction, oral administration of a medicament was generally recognised as being the obvious goal to aim for since it would overcome the unpleasant and potentially hazardous procedures associated with intracavernosal injections. In fact, at the NIH consensus meeting in 1992 it was stated that amongst the needs and directions for future research was the development of new therapies, including pharmacologic agents, and with the emphasis on oral agents that may address the cause of male erectile dysfunction which greater specificity (page 194)". (Pryor Expert Report paragraph 5.4)

(The NIH referred to in that passage is the American National Institute of Health which held a meeting on impotence in December 1992.)

102. I accept Mr Pryor's evidence as fair and reliable. Indeed Pfizer accepts that at the priority date an effective oral treatment for impotence was known to be desirable "in the abstract". It will be seen, therefore, that claim 10 covers the use of any cGMP PDE inhibitor in achieving a well recognised goal, namely the oral treatment of MED. Before addressing the factual question of whether or not it was obvious to try to use such inhibitors for that type of treatment, it appears that a point of principle arises in relation to this claim and claim 11. Assuming that it was obvious to use a cGMP PDE inhibitor or a cGMP PDE_V inhibitor to treat male impotence, is it possible for Pfizer to obtain valid patent protection where the only distinguishing feature between what is obvious and what is claimed consists of the fact that the treatment is orally administered? In addressing this issue it should be borne in mind that it is not suggested that all cGMP PDE inhibitors or cGMP PDE_V inhibitors are capable of being administered orally or that the patent includes any teaching which would allow a reader to predict which of the numerous products it covers would be capable of oral administration. It does not teach any general method of designing chemicals to be capable of oral administration nor does it suggest that any such general method exists. The patentee has found one inhibitor (the identity of which it does not disclose) which is said to be efficacious and which happens to be capable of oral administration. As far as the teaching of the patent is concerned, the only way of discovering whether any inhibitor is efficacious orally is to try it out.
103. In Bourns Inc. v. Raychem Corporation (Judgment 12 June 1997) I said:

"… What has to be determined is what technical contribution to the art has been made by the patentee. If that contribution is obvious then it is not protectable under patent law. If the patent claim consists of no more than a product or process selected by reference to a set of obviously desirable parameters, then the technical contribution is the selection of those parameters. Since that selection is obvious, so is the claim. It is permissible to look at the teaching in the specification to see what the patentee has put forward as his technical contribution. Where, as Mr Silverleaf argues is the case here, the teaching indicates that nothing novel by way of materials or processing has been used, it reinforces the conclusion that the patentee has done no more than select the obviously useful products out of the range of those which can be made with existing technology. In such a case, the patent is just for any good product. On the other hand, where the invention involves the use of new materials or a new process, such as a new way of using known materials, to achieve a known or obvious goal, the inventive concept (per Windsurfing) or technical contribution (per AGREVO) is the materials or process. If the materials or process are not obvious, a claim of permissible width directed to or dependent on the materials or process is not obvious either. Although the claims will give protection to products or processes which meet obvious desiderata, it is the materials or methods for getting there which supports that protection. Here also, the teaching in the specification will be directed at the new materials or processes and will reinforce the conclusion that the claims are directed to a protectable technical contribution."

104. The decision in that case was upheld in the Court of Appeal and, in particular, the findings that a number of the claims in issue were merely to known desiderata were cited with approval. I think the same point applies here. If it was obvious to use a cGMP PDE inhibitor to treat male impotence, then a claim to that could not be valid. Nor could a claim which sought to protect that treatment when administered in an obviously desirable way. If the use of cGMP PDE inhibitors generally or cGMP PDE_V inhibitors specifically in the treatment of MED was obvious, Pfizer have made no technical contribution by testing one or more of such agents and recording that some of them are capable of being administered orally. These claims only cover an obvious desideratum and would fail accordingly.
105. However, there are additional reasons why Mr Kitchin's arguments must be rejected. On the facts here there was overwhelming pressure on workers in the field to try oral administration with any new pharmaceutical candidate for MED treatment. I have already referred to the very unpleasant nature of the primary treatment for male impotence, namely self injection into the penis immediately before sexual intercourse, and the general appreciation by 1992 if not before that oral treatment was what was needed. Pfizer suggests that the perceived risks of oral administration of a PDE inhibitor were so great that they would not even be tried. This argument reached its high point in Professor Ignarro's warning that oral administration risked killing the patients. Because some PDE inhibitors were known to have an effect on the peripheral vascular system which could result in lowered blood pressure, so any systemic treatment would be likely to be potentially life-threatening.
106. It is all too easy in a case like this to raise a lurid fear and for that to obscure the issues. That, in my view, is what Pfizer have done here. It is well known and not in dispute that regulatory authorities exercise immense care before they will license drugs for human use. The prospect of an orally administered impotence treatment being allowed on the market which has a significant risk of killing patients or even causing them significant harm must be regarded as very small indeed. As Mr Kitchin is anxious to point out, most drugs go through extensive testing, including testing on live animals, before they are allowed even to be clinically tested on humans. The risk, if there is one, therefore is not a risk of killing patients so much as of failing to make it through the trials which always have to be undertaken to prove efficacy and lack of relevant or unacceptable toxicity and side effects. The question is, therefore, whether it would be obvious to try to use a cGMP PDE inhibitor in oral treatment or were the risks of failing so great as to deter the notional skilled worker before he set off down that path. Whether something is obvious to try depends to a large extent on balancing the expected rewards if there is success against the size of the risk of failure. Here it was apparent that the rewards for finding an oral treatment would be substantial. The risk was not, as indicated above, the risk of killing anyone, but the risk that trying oral administration would not work so that the research would be unproductive. In considering this, it is worth bearing in mind the approach adopted by the EPO Technical Board of Appeal in case T0379/96, a case concerned with attempts to replace ozone damaging aerosol propellants with non-damaging ones. The Board said:

"Moreover, having regard to the degree of pressure put on industry by existing or imminent legislation and by the public interest, to try to replace P12 [i.e. a damaging propellant], in the Board's view, it is a minor matter whether or not there was a particularly high degree of expected success before starting experimental work with HFC 134a."

107. I have come to the conclusion that the skilled team would not have been put off trying oral administration of a PDE inhibitor. On the contrary, on balance there is much in the evidence which suggests that trying oral administration was a worthwhile, and perhaps the first, avenue to pursue. The skilled team would have included PDE expertise. It would have known that the oral administration of PDE inhibitors was already being tried in a number of fields. As Dr. Challis agreed, it was well known by 1993 that a number of PDE inhibitors were bioavailable orally. Further Lilly produced a document (X17) which referred to 25 patents to PDE inhibitors applied for before the priority date of the patent in suit. The patentees included Warner-Lambert Co., SmithKline Beecham, Pfizer and Schering – all well known names in the pharmaceutical field. Every one of them suggested that the inhibitors should be administered orally. In some cases they contain a suggested daily dose for human ingestion. Even by the time of Rajfer in January 1992, it would have been apparent that it was worth trying oral administration of any candidate compound. The reasons for this are not difficult to understand. On the one side there was the pressure to find an oral treatment for MED. On the other there was nothing to suggest that these inhibitors as a class were necessarily ineffective or toxic when so administered. Dr. Gristwood gave evidence on this issue. He said:

"Having read Rajfer et. al., there was nothing to deter me from pursuing an oral delivery of a selective PDE V inhibitor as a treatment of impotence. Although cardiovascular effects have been reported in anaesthetised rats and dogs with intravenous administration of zaprinast, I was not then, and still am not, aware of any reported adverse side effects associated with oral administration of zaprinast to humans. I was aware in 1993 that zaprinast had been administered orally in humans (both adults and children) without adverse side effects. I would expect most people working in the field to have been aware of this."

108. I accept that evidence. The fact that zaprinast (May & Baker's PDE_V inhibitor) which was discussed and used in Rajfer was the subject of clinical trials and had been administered orally for that purpose would have been known to all those involved in the potential use of PDE inhibitors as a drug candidate. Anyone who had any fears about the use of it, or other PDE inhibitors, for oral administration would have found that out from the published literature. Furthermore, even if the skilled man was not aware that PDE inhibitors had actually been used in such clinical trials, he would not have been put off trying oral administration. He would have been unable to predict whether there would be a problem associated with oral administration, or if there was its nature and magnitude and he would be reassured that if any significant problem did exist it would surface during the course of the tests necessary for regulatory approval, that is to say tests specifically designed as a filter to ensure no appreciable risk to the public. The result is that the skilled man or team in the art would not have been put off from trying to use a PDE inhibitor for oral administration.
109. The evidence supports the view that any worries about possible side effect or contra indications would not have been sufficient to prevent a worker in the field from trying out PDEs for administration to man, whether orally or otherwise. For example Dr. Challis' evidence was:

"Q. Coming back again to paragraph 46 you are unable to point my Lord to any known PDE5 as at 1993 which was known to have life-threatening side-effects.

A. I am sorry, I have gone to the .... We are on 46?

Q. Did I say 46. I should not have done. …

MR. THORLEY: I do apologize; it is my fault. Paragraph 65, top of page 29, you were talking about life-threatening consequences. That is what we were talking about and I am trying to ensure that I have got your evidence accurately. Do you want me to put the question again?

A. No. I am quite happy. I am saying that, I think the point that I am trying to get across there is that if one is thinking about the treatment of MED one has a good idea of what the patient population is going to be. It is going to be generally more elderly. I think the incidence of impotence increases essentially exponentially beyond the age of around 40, so we are dealing here with a group of patients that have potentially other health problems in their lives. We are dealing with a disorder that is in itself not life-threatening and, therefore, I think that anyone developing drugs in this area is going to be acutely aware of any side-effects, never mind life-threatening side-effects, but there is the potential of these drugs, I think, to cause life-threatening effects in people with particular other complicating disorders.

Q. Are you saying that that concern would deter you from doing any development work in the field or are you saying that you would conduct development research with caution?

A. The latter."

(Day 5 Transcript page 700)

110. He returned to the same theme on the next day:

"MR. THORLEY: … doctor, you said, if I recall rightly that because of these complications there could be no guarantee that a beneficial therapy could be produced.

A. No.

Q. By pursuing any particular route of potential treatment.

A. That is absolutely the case, yes.

Q. That is the point that you are seeking to make.

A. The point is that I believe .... The sildenafil example gives us a clear example of where, despite all of my provisos, a drug is extremely effective. Therefore, all I am doing is trying to make you aware of the complexities of the field. I am not saying that there was no point in travelling towards that therapeutic goal at all." (Day 6 Transcript page 720)

111. In addition to these points, I should refer to the evidence of Dr. Dennis Smith, the Senior Director of Pfizer's Drug Metabolism Department and whose entire industrial career has been in the area of Drug Metabolism and Pharmacokinetics. He was asked by Pfizer to comment on what procedures would have formed a routine part of drug development programmes in order to assess the oral bioavailability of a compound, as at June 1993 . In particular, he was asked to comment on the use of conscious animals for this purpose as at that date.
112. He stated that oral bioavailability assessment of compounds using animals, such as the dog had certainly become a well known and routine part of drug development programmes by 1992. He said that this was due to fundamental similarities between animals and man in the processes governing absorption of compounds from the gastrointestinal tract and subsequent appearance in the systemic circulation (from where the drug molecule exerts its pharmacological effect). He also said that such assessments are a routine aspect of selecting the appropriate drug molecules for clinical development. He supported that view by a reference to numerous published articles from various workers, from a large number of pharmaceutical companies. His report also contains the following:

"I should say that in practice, oral bioavailability studies are not carried out on the full range of compounds studied in the course of a drug discovery programme, but only those which are considered to be the more promising candidates for further development. For each compound being tested, such studies normally took approximately two to three weeks to carry out in 1993. I should also point out that species differences in bioavailability between man and animal species, including dog, have been identified, which may mean that the data needs interpretation for prediction to man. However, the assessment of oral bioavailability in animal species in 1993 and today remains an essential element of the preclinical assessment of novel drug candidates. Increasingly in vitro systems have been put into place which facilitate this interpretation, but these are used alongside data on bioavailability in conscious animals. Other animal species are used in addition to the dog in the assessment of oral bioavailability. These include mice, rats and monkeys. However, the extensive use of the dog in pharmacology studies and in drug safety evaluation within the pharmaceutical industry, results in the dog being amongst the most commonly used of the animal species."

113. Unsurprisingly, Lilly chose not to cross examine this witness. His evidence supports the view that it would have been a matter of course to try out any potential anti-MED candidate by oral administration to conscious animals, including dogs, and that any such testing would have taken a comparatively short time. That is what the skilled man in the art would have done.

(f) Others, including Pfizer, did not hit on the idea and commercial success

114. Mr Kitchin also argues that allegations of obviousness levelled at all of the claims are inconsistent with the fact that others in the pharmaceutical industry, including initially Pfizer, missed the idea of using cGMP PDE inhibitors for the treatment of MED and that the introduction of Viagra has met with marked commercial success.
115. The suggestion that others in the art missed the invention, even if true, is a mirror of Lilly's argument that many in the art read Rajfer as teaching the use of PDE inhibitors for the treatment of MED. It should be treated with caution for much the same reasons as have been set out at paragraphs 63, 64 and 75 above. Furthermore such initial impact as it may have is diminished by the suggestion, much emphasised by Pfizer, that there were many other obvious courses which those in the art could follow. Thirdly, the period between the publication of Rajfer and the priority date of the patent is only 18 months. The fact, assuming it to be so, that no one else had put in hand a project for trying to use PDE inhibitors for treating MED in that short period is little indication that the idea is inventive. This also ties in with Mr Kitchin's argument on commercial success. This is a topic discussed at length in Haberman v. Jackel [1999] FSR 683. It would serve no purpose to revisit that issue at length here. Even doing what is obvious can be commercially successful. Commercial success comes into its own as a secondary indication of inventiveness where both the relevant prior art has been available and the need for a solution to a known problem has been sought for a long time. Failure to make the step which is covered by the patent in those circumstances may be some indication that it is not as obvious as it might at first appear. That has no application here where the gap between the prior art and the priority date is so very short.

Conclusion in relation to Rajfer

116. In my view, Mr Thorley's arguments on obviousness prevail against all the claims in the patent. The man in the art at the priority date would have been aware of the excitement surrounding the discovery of nitric oxide's role as a messenger for the relaxation of smooth muscle. He would have appreciated that Rajfer, probably for the first time, had elucidated the whole path. He would have understood the teaching that the creation and maintenance of elevated levels of cGMP was an important factor in the relaxation of corpus cavernosum muscle. In the light of this Rajfer's suggestion that interference with this pathway might cause some forms of impotence and that it could be treatable would be seen as a sensible proposal worth taking forward. Such a man skilled in the art would have appreciated that there were at least two ways of producing and maintaining elevated cGMP levels, namely to produce more or to destroy less. That is what Rajfer's experiments and paper teach. It would have been obvious in these circumstances that it would be worth trying to utilise either or both ends of the pathway to treat patients with MED, including those who had inadequate or short-lived erections. Further Rajfer indicates that a known medical agent, zaprinast, works on the back end of the pathway and its mode of action, namely that it was a cGMP PDE inhibitor, and indeed a selective one, was explained. This would have made it obvious to try not just zaprinast but other cGMP PDE inhibitors. For reasons set out above this would have led to Bell I. Similarly, for reasons set out above, it would have been obvious to try oral administration.
117. The nub of this analysis can be expressed more simply. Rajfer teaches that in vitro zaprinast enhances the relaxation of the muscles responsible for causing an erection. It is likely that the same effect could be achieved in vivo. Further, Rajfer explains how zaprinast was believed to work, namely because it is a cGMP PDE inhibitor and this would have lead a skilled worker to try other such inhibitors using any convenient and attractive form of administration. Oral administration was the most obvious to try.

(B) Murray

118. Although I have found that Rajfer renders the patent invalid for obviousness, Lilly argues that an entirely separate and stronger case of obviousness can be made out based on the Murray paper. Because much of the evidence went to the disclosure in Murray and because this case is likely to go further, it is appropriate to consider this as well.
119. Murray is a review article. It identifies its author as the Senior Biologist in the Cyclic Nucleotide Research Programme of SmithKline Beecham Pharmaceuticals in England. There was no suggestion before me that it would be considered as anything other than a serious and respectable publication. It was published a month or so before the priority date of Pfizer's patent. As a review article, it makes reference to a number of pieces of prior published material. It is not in dispute between the parties that it would have been read carefully. Mr Thorley said that it would have been read in conjunction with the documents to which it specifically refers. This would include, in particular, Rajfer. Mr Kitchin accepted, as I understand it, that Murray would be read carefully together with some of the documents to which it referred. He expressly accepted that someone reading Murray would look back and read Rajfer, were that person looking at Murray with any interest in MED. For the reasons set out below, I have come to the conclusion that a person skilled in the art would read Murray with an interest in MED, so it and Rajfer should be considered together. Mr Kitchin went further and said that some of the other references in Murray should also be taken into account when deciding what Murray teaches. I accept that. At the very least the reader would carefully read the list of references at the end of the article to decide which of them were worth consulting further. As a result he would obtain copies of all of them save those which would clearly be treated as of peripheral interest only.
120. Murray indicates the focus of his article with its title, "Phosphodiesterase V_A Inhibitors". Although the document should be read as a whole, there are some paragraphs which are particularly important as far as the issue of obviousness is concerned. These are set out below. For ease of subsequent reference, each is identified by number. The first paragraph reads as follows:

Extract 1: "Cyclic nucleotide phosphodiesterases (PDEs) have long been regarded as potential targets for therapeutic agents. More recently, interest in this area has been renewed by the recognition that there are five distinct PDE isoenzyme families, and that tissues have different complements of these isoenzymes. There is, therefore, a logical foundation for selective PDE inhibitors to be used to increase cyclic nucleotide levels in specific target tissues or organs. Indeed selective PDE III inhibitors (see below for nomenclature) are currently used clinically for the treatment of congestive heart failure and as antithrombotic agents, and this success has given hope that inhibitors of other PDE isoenzymes will also be useful drugs. The purpose of this review is to discuss briefly the PDE isoenzyme families and to describe one of these isoenzyme families, PDE V. The physiological and pharmacological effects of PDE V_A inhibitors are reviewed and their potential therapeutic indications discussed."

121. It goes on to discuss the PDE families and nomenclature and then describes PDE V_A as a member of a cGMP-specific PDE isoenzyme family. It discusses which tissues PDE V_A is found in and continues:

Extract 2: "However, when compared to some other PDEs (e.g., PDE IV), PDE V_A exhibits a rather limited tissue distribution. … Probably the most salient feature of PDE V_A is its specificity for cGMP as a substrate, and, among the PDE isoenzymes, this is a unique property of PDE V_A. This means that inhibition of PDE V_A will result in elevation of cGMP, but not cAMP levels … In contrast, PDE IV inhibitors cause an elevation of cAMP, but not cGMP, while PDE III inhibitors can increase the levels of both cyclic nucleotides. The combination of a limited tissue distribution and substrate specificity suggests that a specific PDE V_A inhibitor could have a narrow range of physiological and pharmacological actions."

122. Murray records that the most frequently studied PDE V_A inhibitor is zaprinast and, by reference to a table, indicates that there were by this time at least three selective inhibitors of PDE_V, namely zaprinast (a May & Baker product), MY-5445 and SK&F 96231. The latter which, because of its name, was clearly a product of SmithKline Beecham, is described by Murray as an expressly designed prototype PDE V_A inhibitor. In discussing zaprinast, Murray says that there is no doubt that it is a potent inhibitor of PDE V_A. He also notes that PDE V_A is also inhibited by nonselective PDE inhibitors such as papaverine. On the fourth page, he says:

Extract 3: "The effects of zaprinast have also been studied in a number of other smooth muscle types. With strips of human corpus cavernosum, zaprinast alone caused a relaxation and enhanced the relaxation caused by nitric oxide or electrical stimulation."

123. And then:

Extract 4: "Zaprinast is a weak relaxant of lower esophageal sphincter muscle from a number of species, including humans, but potently relaxes guinea pig colon and Taenia coli preparations. Contracted rat gastric fundus is also relaxed by zaprinast, and the compound potentiates the sodium nitroprusside (SNP)-induced relaxation of this tissue.

In vivo studies of PDE V_A inhibitors have largely been on the effects of zaprinast on mean arterial blood pressure (MABP) in anesthetized rats or dogs, and again, the results give a reasonably consistent picture. Zaprinast lowered MABP mainly by decreasing total peripheral resistance, and there is evidence for selective vasodilatation. Zaprinast increased cardiac output and also caused a marked natriuresis, but had little effect on heart rate. In a rat aortavenocaval fistula model of cardiac failure, infusion of zaprinast caused natriuresis with little effect on MABP. Treatment of spontaneously hypertensive rats with zaprinast caused a fall in MABP to control levels. SK&F-96231 did not affect heart rate or MABP in conscious dogs, whereas bronchodilating effects of the compound were observed in anesthetized guinea pigs."

124. The article ends with three important paragraphs under the rubric "Potential therapeutic use of PDE V_A inhibitors":

Extract 5: "To date, the only PDE V_A inhibitor that has been clinically evaluated is zaprinast. In adult asthmatics, zaprinast reduced bronchoconstriction induced by exercise, but not that provoked by histamine⁵¹. In contrast, zaprinast had no effect on exercise-induced asthma in children⁵². However, when evaluating these results, it should be remembered that these were small clinical trials using a compound that may have actions other than PDE V_A inhibition. Therefore, at present, PDE V_A inhibitors must be considered to be compounds with potential rather than established clinical use."

And

Extract 6: "Smooth muscle relaxation appears to be the most promising of the potential uses of PDE V_A inhibitors, and possible therapeutic utilities could include vaso dilatation, bronchodilatation, modulation of gastrointestinal motility and treatment of impotence. Although PDE V_A inhibitors will relax most smooth muscle types in vitro, this does not necessarily mean that a nonselective action will be seen in vivo. Selective actions could be achieved by virtue of the fact that many of the effects of PDE V_A inhibitors in a particular tissue are dependent of the level of guanylate cyclase activity. Thus, PDE V_A inhibitors will have the greatest effect in cells and tissues that have a high guanylate cyclase activity, and there could be considerable value in a therapeutic agent that has little activity in its own right, but potentiates the effects of endogenous mediators. An example of this has been demonstrated in the rat model of cardiac failure described above, where the operated animals were more sensitive to the effects of zaprinast than their sham-operated controls. An explanation for this is that higher levels of the guanylate cylcase-activating ANP are found in the former group."

and

Extract 7: "… At present the therapeutic potential of PDE V_A is largely based on the effects of … zaprinast, and a clearer picture will be obtained when other rationally designed PDE V_A inhibitors become available."

The article ends with a list of references.

125. In view of the concession that it is permissible to read this article together with Rajfer, it must follow from what has been set out above that it also renders Pfizer obvious. However Murray is an even stronger platform upon which to base Lilly's arguments. Murray concentrates on the back end of the L-arginine-Nitric Oxide-Cyclic GMP pathway, encourages the reader to consider using selective PDE_V inhibitors for the therapeutic treatment of, inter alia, impotence and makes it clear that although a number of such selective inhibitors exist, more are likely to become available. It seems to me that this document would directly encourage the skilled addressee to do what Pfizer has covered in its patent. The strength of this teaching and the weakness of the arguments against it can be illustrated conveniently by reference to the heads of Mr Kitchin's arguments set out above.

(a) Alternative routes to a treatment

Whatever the position in relation to Rajfer, Murray directs the reader to think in terms of therapeutic treatments of impotence using cGMP PDE inhibitors and, in particular, PDE_V inhibitors. There is no question of the reader dismissing this teaching as obviously untenable. It was a serious suggestion made by someone closely involved in PDE research for a leading pharmaceutical company. Even if other routes were to occur to him, the notional skilled addressee would read this as directing him to use inhibitors and it would be obvious and natural for him to follow that suggestion.

(b) Only the front end of the NANC pathway would have been of interest

126. In relation to Rajfer, Mr Kitchin is able to argue that the skilled addressee would concentrate on part of the teaching, namely that relating to the front end of the NANC pathway, to such an extent that there would be no substantial consideration of the back end of the pathway. For reasons set out above, I do not accept that argument. However it is close to impossible to run it in relation to Murray. As its title and all of its teaching makes clear, the article is concentrating on what PDE inhibitors do and can be used for. It should be noted that the first sentence in extract 6 points to the "most promising" uses of PDE V_A inhibitors and specifically refers to potential therapeutic use in the treatment of impotence. The whole flavour and direction of the paper is summarised in the last sentence of extract 1 namely:

"[In this paper] the physiological and pharmacological effects of PDE V_A inhibitors are reviewed and their potential therapeutic indications discussed."

127. I should mention that Professor Ignarro said that the Murray paper would give him "no incentive to investigate the use of PDE V_A inhibitors as a treatment for impotence, and certainly no reason to deviate from the nitric oxide donor approach." (see Bundle G1, Tab 3 paragraph 34). I have no reason to doubt the sincerity of the Professor's views but it seems to me that this part of his testimony is an illustration of his pre-occupation with the front end of the NANC pathway. This is something I will consider more fully in relation to the Bush paper. However that preoccupation would not be shared by an ordinary unimaginative skilled man reading this document at the priority date. On the contrary, the skilled worker would read with particular interest the statements in Murray, notably in extract 2, that PDE V_A has limited tissue distribution and that its inhibitors are likely to have a narrow range of physiological and pharmacological actions. This would suggest that these inhibitors could be targeted quite narrowly on to specific tissues or organs. By comparison the nitric oxide end of the pathway will be found in all tissues. This would reinforce the attractiveness of the Murray recommendation.
128. The article was put to Dr. Challis. His evidence was as follows:.

"Q. Thank you. Can we now look at the article. Can I just, to make the position clear, ask you to look at the left-hand side of the first page, starting at the beginning: "Cyclic nucleotide phosphodiesterases have long been regarded as potential targets for therapeutic agents. More recently, interest in this area has been renewed by the recognition that there are five distinct PDE isoenzyme families, and that tissues have different complements of these isoenzymes. There is, therefore, a logical foundation for selective PDE inhibitors to be used to increase cyclic nucleotide levels in specific target tissues or organs." Is that fair comment?

A. Perfectly fair, yes.

…

Q. He has done his best to communicate his views to the public in this article.

A. I think he, as you are, is constructing a case here. He is making a case for why it is important for people interested in this area to examine PDE5 and the inhibition of PDE5 as a therapeutically potentially beneficial target.

Q. He is putting forward the justification that he feels is there for approaching that task.

A. He is, yes.". (Day 6, Transcript page 710)

129. I did not understand Dr Challis to be criticizing Dr. Murray when he said that he was "constructing a case". It was apparent that Dr Challis had considerable admiration for Dr Murray's paper. What Dr Challis was saying, accurately in my view, was that the article reads as if Dr Murray was trying to persuade the reader that his views and recommendations were sensible. The skilled man in the art would have been persuaded.

(c) Only a combined treatment would be thought of

130. Once again, it appears to me that it is very difficult indeed to put this argument forward in relation to the Murray paper. The suggestions in the paper relating to possible therapeutic applications concentrates on PDE_V inhibitors alone. This does not exclude the possibility of using such inhibitors in conjunction with other agents, but it does suggest that the inhibitors may well be useful by themselves. That message is reinforced by extract 3 which cross refers to the Rajfer paper. Murray clearly had read the latter paper with care and he tells his reader that zaprinast alone had caused relaxation of corpus cavernosum smooth muscle. It is not in dispute that Murray's deduction is correct. The experimental data set out in Rajfer does show that zaprinast was effective when used in vivo alone. Again, this was the subject of comment by Dr. Challis under cross examination. He accepted that Dr Murray was entitled to draw the conclusion that Rajfer had demonstrated that zaprinast alone caused a relaxation of the smooth muscles in the penis. He complemented Dr. Murray on the care with which he had read the Rajfer paper.

(d) No reason to find the Bell applications

131. Again, it seems to me that this argument is harder to advance in relation to Murray. Any skilled reader would notice that this is a paper written by a senior researcher in one of the largest pharmaceutical companies, SmithKline Beecham. It is apparent that this is an area of research in which not only that company but other pharmaceutical companies are engaged. Thus it is clear that zaprinast is a May & Baker product. Extract 1 indicates that this is an area of current interest in the industry. As I have said in relation to Rajfer, a man skilled in the art would naturally search among the patents of other pharmaceutical companies to find candidate PDE inhibitors and doing so would inevitably turn up Bell I. The teaching in Murray would encourage him to look in this direction.

(e) Oral administration would be shunned

132. Here also the teaching of Murray undermines Mr Kitchin's argument. It will be remembered that Mr Kitchin suggests that the man in the art would think that the side effects of administration of a PDE inhibitor were likely to be so dangerous that it would not be worth trying to use it as a drug. The particular concern was that because PDE inhibition, or PDE_V inhibition, might relax the smooth muscles throughout the vascular system, this could lead to strong adverse affects, for example on blood pressure. This fear would be particularly heightened if the inhibitor was administered in large doses and was made available systemically – e.g. if it was administered orally. Hence the reference to 'killing patients'.
133. If the skilled addressee had such concerns (and for reasons already given, I do not think that any concerns he might have would have deterred him from trying to use a PDE inhibitor orally), Murray goes a long way towards setting them aside. This can be analysed in two parts. First consider the suggestion that a PDE, or a PDE_V inhibitor, would have been thought to be so dangerous that it would not be worth trying by any mode of application. This goes to claim 1 and its dependents. For any such suggestion to prevail it would necessitate ignoring virtually the whole of the teaching of Murray. As noted already, Dr. Murray was identified in his paper as the Senior Biologist in the Cyclic Nucleotide Research Programme of one of the best known research based pharmaceutical companies in the world. The major teaching of his paper is that PDE V_A inhibitors should be tried as therapeutic agents in the treatment of human diseases. As noted above, extract 5 is preceded by the rubric; "Potential therapeutic use of PDE V_A inhibitors" and the following passage identifies four suitable candidate diseases which might be susceptible to treatment by this class of pharmacologically active agents. If a man in the art would be likely to think that such inhibitors were too dangerous to be tried for therapeutic use in man, it would mean that the whole of the Murray proposal was misconceived and that this senior researcher was recommending the trial of something which was obviously too dangerous to administer. Worse than that, Murray expressly refers to the vascular effects of these agents (see extract 4), yet a man in the art would conclude that the author had missed the overriding dangers which would have made any such attempted use futile.
134. Furthermore, Murray is not alone. The article starts by referring to the fact that PDEs have long been regarded as potential targets for therapeutic agents and that that interest has now been increased by the knowledge that they are quite tissue specific (see extracts 1 and 2). On the second page of the article there is a table which sets out known PDE inhibitors. They include not only zaprinast but also milrinone and rolipram. There does not appear to be any dispute that a man in the art would have known that they had been tried out on man (see for example, Mr Kitchin's cross examination of Dr. Kruse on Day 4, transcript page 478). In addition, on page 4 of the article, Dr Murray sets out the structural formulae of a number of known PDE_VA inhibitors. They include papaverine, the most extensively used anti impotence drug in the United Kingdom. It must follow that pharmaceutical companies did not share the belief now put forward by Pfizer that these inhibitors were so dangerous that they could not be tried. Pfizer's argument on this point is untenable. Once again it is instructive to consider Dr. Challis' evidence:

"Q. The teaching that is coming out of this is, "Go and look for your own PDE5. There are ones that can be made that are better than the old workhorse zaprinast".

A. I think that is what the masthead of the paper is essentially saying, so it is there, yes.

Q. There is no suggestion in this paper that the potential side-effects, either of zaprinast or of the SK&F product, are in any way life threatening?

A. I think he is flagging up the side-effects in a number of animal models. If we go back to life threatening, I think a healthy anaesthetized rat, and looking at the effects in there, you can perhaps trivialize some of the effects but they may be important in the target population of humans.

Q. The purpose of him giving this information is to encourage research into PDE5 inhibitors by saying to them, "You may not find that there are side-effects preventing the use of the products".

A. He certainly wishes to get across the message that any side-effects problems hopefully are not unsurmountable." (Day 6, Transcript page 712)

135. Furthermore Murray is difficult to reconcile with Mr Kitchin's argument even if it is limited to a concern not to use these inhibitors for oral treatment. Any skilled addressee reading the Murray paper carefully and concerned about the potential adverse effects of these agents would have noted the author's encouragement to use them because of the limited tissue distribution of PDE_V. I have referred to this already. But more than this, if the skilled addressee really had such concerns he would pay particular attention to that passage in the paper which touches upon the fact that at least one PDE V_A inhibitor, zaprinast, had already been used in clinical trials. This is the passage set out in extract 5 above. The place that clinical trials play in the development of a potential drug are described in the evidence of Dr Kruse, one of Lilly's witnesses. Suffice to say, no drug would be the subject of clinical trials unless there were strong reasons to believe that it would not be seriously dangerous to humans. So, extract 5 points out that zaprinast had already been clinically evaluated. More than that it refers to the fact that such clinical evaluation had been in adult asthmatics and also in children suffering from exercise-induced asthma. Each of these claims, the accuracy of which has not been disputed, is backed up by references to papers. They are identified in extract 5 as footnotes 51 and 52. The titles to those papers, as set out in Murray, are respectively:

"Rudd, R.M., Gellert, A.R., Studdy, P.R. and Geddes, D.M. Inhibition of exercise-induced asthma by an orally absorbed mast cell stabilizer (M&B 22,948)"

and

"Reiser, J., Yeang, Y. and Warner, J.O. The effect of zaprinast (M&B 22,948, an orally absorbed mast cell stabilizer) on exercise-induced asthma in children."

136. In my view it is most unlikely, in the light of those references, that the man skilled in the art would be deterred from trying oral administration of potential PDE inhibitors. On the contrary, the fact that zaprinast had been considered safe enough to be used orally to treat patients, including children, would have tended to encourage him into thinking that oral administration of inhibitors was a real possibility. Mr Kitchin suggested the fear would still exist because very high blood levels of zaprinast might be expected to be necessary to treat impotence. However this really takes the case no further. Whether very high blood levels of a suitable inhibitor would be necessary and, if so, the adverse side effects of it are not matters which can be predicted in advance. The skilled man would not have been deterred by these considerations either.

(f) Others, including Pfizer, did not hit on the idea and commercial success

137. Whatever the position in relation to Rajfer, it seems that Murray did hit on the essential ideas behind the Pfizer patent and published them so this point does not run. Furthermore the argument of commercial success does not run either. It is Pfizers' position that Rajfer did not render the claims of its patent obvious. If so, the starting point for the obviousness attack is Murray. The argument of commercial success cannot succeed where there is only a matter of a few weeks between the pleaded prior art and the priority date.

Conclusion in relation to Murray

138. I accept Mr Thorley's arguments of obviousness. If anything, the case of obviousness based on Murray is even stronger than in relation to Rajfer. It could be said that Pfizer's patent is little more than a putting into practice of the recommendations and suggestions of Dr Murray. Those recommendations and suggestions would have been appreciated by a man skilled in the art at the priority date as being sound and worth trying.

(C) Bush

139. Dr Bush's thesis was published after Rajfer and refers to it. It was written before Murray had been published. Dr Bush was a junior member of Professor Ignarro's team. Her thesis had to be 'defended' by her in an oral examination before she could be awarded her doctorate. The examination panel consisted of Professor Ignarro, Professor Gautam Chaudhuri, who held a joint appointment in pharmacology as well as obstetrics and gynaecology, Dr. John Fukuto, an Associate Professor in pharmacology, Dr. Joseph Gambone, a Professor in obstetrics and gynaecology and Dr. Rajfer. They all signed their approval on the second page of the thesis and it is the thesis bearing their signatures which is conceded, for the purpose of this action, to have been published before the priority date. No one has suggested that its contents could be dismissed as trivial, obviously wrong or illogical. It is 179 pages long and describes the anatomy of the penis, the physiology and pharmacology of erection, impotence, the NANC pathway and a number of allied topics. Among the agents used for the experiments which underpin the thesis is zaprinast. The disclosure appears to cover everything of significance which was described in the Rajfer paper. To that extent the same arguments of obviousness can be run and apply to this paper as well. However Bush goes further than Rajfer. For present purposes it is only necessary to pay particular attention to chapter 4 which sets out Dr. Bush's summary and conclusions.
140. Whereas Murray concentrates on the back end of the NANC pathway and the potential therapeutic uses of PDE inhibitors, Bush, like Rajfer, looks at both the front and back ends. At page 157 she states that there is now strong evidence to support the role of nitric oxide as a neurotransmitter in the corpus cavernosum. She goes on to note:

Extract 1: "Anecdotal reports of enhanced erection following administration of various nitrovasodilators can now be better understood, in the light of the involvement of nitric oxide in penile erection. One case report in the literature described an experience of a previously impotent man who received nitroglycerin for chest pain. The individual experienced his first erection in several years after taking the nitrate, and subsequent erections, some of which were painful, occurred several times a week and were temporally associated with administration of the nitroglycerin. There is documentation of abuse of volatile nitrites such as isoamyl nitrite over the past three decades for their aphrodisiac properties. The colloquial term for amyl nitrite is "popper" because the drug is available in a glass ampule (encased in a fabric covering) which is crushed just prior to inhalation, resulting in a "popping" sound. These agents have been widely abused for the purpose of enhancing sexual performance and pleasure." .

141. There is no dispute as to the meaning of this passage. The reader of this passage would understand that nitroglycerin and volatile nitrites such as isoamyl nitrite release NO. It is for this reason that they are called nitrovasodilators. Dr. Bush was explaining that the side effect of enhanced sexual activity and, in particular, erections associated with the application of these NO producers was likely to be due to the fact that release of NO in the penis results in relaxation of the smooth muscle in the corpus cavernosum. As many non-medical members of the public are aware, nitroglycerin is normally administered by placing a tablet under the tongue of the patient. The active ingredient is then absorbed into the bloodstream through the skin surface, or epithelium, in the mouth. This is buccal administration. As Dr. Bush explains, amyl nitrite in 'poppers' is administered by inhalation. In that case also, the active ingredient passes through a membrane (in this case in the lungs) and into the blood. Thus in both cases the active agent is administered in a way which involves it being carried to the penis in the blood stream, that is to say systemically. The more important passage in the thesis for present purposes is on pages 159 and 160. I set out the relevant parts below:

Extract 2: "The results of this research are probably most important in terms of practical application to the treatment of urological disorders such as impotence and priapism. Now that a physiological mechanism for corporal smooth muscle relaxation has been established, this mechanism can be used as a framework to systematically study the problem of impotence. The etiology of vasculogenic and/or neurogenic impotence may be linked to a defect somewhere in the L-arginine-nitric oxide-cyclic GMP pathway. Expanded knowledge of the physiological mechanism of erection also allows for the design of rational drug therapy for the treatment of disorders such as impotence and priapism. For example the selection of nitrovaosdilators for intracavernosal injection or by alternative drug delivery forms, is a very rational treatment, since these would mimic the natural physiological process. Knowledge of the factors regulating corporal smooth muscle relaxation can be used to critically evaluate the current treatment options for impotence and priapism. For example, some of the agents currently being used for the treatment of impotence and priapism appear to be good choices from a physiological basis. … Papaverine probably acts mainly through phosphodiesterase inhibition and more specifically through inhibition of [cGMP PDE]. Clinical development of a specific [cGMP PDE] inhibitor should be considered for the treatment of impotence. A specific [cGMP PDE ] inhibitor could enhance corporal smooth muscle relaxation and produce erection by inhibiting the breakdown of cyclic GMP, thus having a direct and specific effect on the L-arginine-nitric oxide-cyclic GMP mediated relaxation process. Agents that are currently being used to treat impotence, the mechanisms of which do not appear to have a physiological basis, should probably be re-evaluated, unless efficacy has been clearly established. …"

142. As with the Murray paper, it seems to me that this makes some of Mr Kitchin's arguments particularly difficult to sustain. I will deal with each of them in turn.

(a) Alternative routes to a treatment

143. The reader of Bush would be directed towards considering therapies based upon the NANC pathway. Bush does not consider any other types of treatment.

(b) Only the front end of the NANC pathway would have been of interest

144. It is difficult to see how this argument can be run in the light of the contents of extract 2. First Bush acknowledges, as a man skilled in the art would have to, that it was not known where the breakdown in the NANC pathway was. From her explanation of how certain nitrovasodilators and papaverine work, it was reasonable to expect that some types of impotence could be cured or ameliorated by interfering with or modifying that pathway, but the place of breakdown was not known. For that reason she says that impotence may be linked to a defect "somewhere" in the pathway. It follows that it would not be possible to predict on the basis of this document that such defects always involve the failure to produce any NO. It may be at the front end or the back end of the pathway or both. Furthermore she explicitly recommends the clinical development of specific cGMP PDE inhibitors. That recommendation could not be dismissed, and was not attacked, as unreasonable or illogical. On the contrary it follows immediately after her suggestion that the efficacy of papaverine, one of the best known treatments for impotence, probably was attributable to its inhibition of cGMP PDE. This was accepted by Professor Ignarro as a 'reasonable conclusion' (Day 3 Transcript page 401). In these circumstances, it is difficult to see why any skilled reader should ignore it.
145. The high point of the objection to Bush's recommendations came from Professor Ignarro. He categorised them as speculation. I do not accept the Professor's views, though undoubtedly firmly and sincerely held, as representing the approach of the skilled non-inventive addressee. Earlier in this judgment I commented on the differences between the notional skilled worker and the idiosyncrasies of real workers in the field. This is illustrated particularly clearly by the evidence given by the Professor. He was and is a man of great intellectual ability. He is also a man who clearly can devote single minded attention to a problem he wants to solve. As he explained, his Nobel Prize was conferred for his overall work in elucidating the properties of nitric oxide as a signalling molecule in the cardiovascular system. It was he and his team who, in the late 1980s, demonstrated that EDRF was nitric oxide. I do not think it would be unfair to think of him as the ultimate expert in the generation of NO as a smooth muscle relaxing agent. However his attitude to the back end of the NANC pathway was clearly conditioned by his personal experience before the publication of the Rajfer, Murray and Bush papers. He put it this way in his report:

"… I had by [1992] been involved for 15 years with various drug companies developing PDE inhibitors for various indications, and all had failed, so I was rather sceptical about their potential as drugs at all."

146. This was the subject of cross examination. The Professor disclosed that he had not been directly involved in the development of drugs but that over a 15 year he had been consulted by a number of companies who had expressed an interest in using PDE inhibitors as potential drugs. None of those consultancies seem to have gone very far. There was no suggestion that any of this consultancy work or any of the programs run by the pharmaceutical companies with which he was involved were public knowledge. However his personal exposure to lack of progress with PDE inhibitors left him with scepticism. It is clear from the evidence at the trial that this scepticism was not shared by many in the pharmaceutical world since many of the largest companies seemed to have embarked on programs to find therapeutic uses for PDE inhibitors. Indeed, the Professor's evidence was also interesting in that he stated that all of the companies for whom he consulted in relation to PDE inhibitors were seeking to use them as orally applied drugs.
147. The area of the Professor's interest was apparent throughout his evidence. This can be illustrated by reference to extracts from his cross-examination. They are just examples:

"Q. That elevation [in cGMP referred to in the Bush thesis] was contributed to by a balance of two factors: one was the presence of enough nitric oxide to create the cGMP in the first place; the other was the absence of an excess of PDE which would hydrolyze the cGMP and so reduce the level.

A. We did not address the latter point. We were focusing on the signal transduction mechanism which means nitric oxide released from the nerves activates guanylate cyclase to stimulate cyclic GMP formation and then the cyclic through a number of different pathways which we did not know about would cause relaxation. Phosphodiesterase inhibition would not really fit into that picture at all. The phosphodiesterase inhibitors, as I mentioned earlier, were used as tools as was methylene blue, hemoglobin, indomethacin, these were all tools that we used to dissect out various parts of the pathway in order to put together the pathway." (Day 3 Transcript page 390)

And:

Q. I think it is Dr. Ellis who gives us the happy analogy of a tap and a plug. Do you understand what I mean by that?

A. Yes; of course, yes.

Q. There is obviously a concern, if you want to fill the basin, that the tap does not turn on at all. Equally, I suggest to you there must be a concern that if your plug hole is so large that no matter how hard you turn the tap on you are never going to get a reservoir of water. Is that fair?

A. I think that is fair as an analogy, yes.

Q. All I am suggesting to you is that it is not illogical to think of the same when considering the nitric oxide pathway?

A. Perhaps it is not illogical. The problem that we are faced with here is we are sitting here in the year 2000 and you are asking us about, you know, or me about what was the thinking in 1993. The thinking now is very different than the thinking back in 1992 or 1993. Back in 1992 and 1993 I was not thinking about that particular concept. I was thinking only about the possible defect in the release of nitric oxide. I was thinking more about, you know, is it possible that one can develop a nitric oxide donor-like molecule that would assist in the treatment of impotence. I was not thinking about the other end. Perhaps I should have -- it may be I was thinking illogically. Maybe I was not smart enough. Nevertheless that was our thinking at that time based on all the available evidence and knowledge and publications and pharmacology urology and patho physiology. (Day 3 page 392)

148. Needless to say, Lilly has never suggested that the Professor was either illogical or not smart. But it does argue that his concentration on one part of the pathway to the exclusion of the other is not an attribute which would be exhibited by the notional skilled worker. The Professor's attitude to the use of PDE inhibitors can also be seen clearly in his report. When commenting on the Bush thesis he said:

"[Bush] goes on to say that "clinical development of a specific cGMP PDE inhibitor should be considered for the treatment of impotence." For the reasons I have already given, I would not have thought of this as the way forward for any sort of treatment of impotence …"

149. I have no doubt that this accurately reflects his views. However the fact that he would not have thought of the idea does not dispose of the fact that Bush had done so and had proposed it in her thesis. The Professor's attitude is a product of his single minded concentration on the front end of the NANC pathway and his lack of enthusiasm for the back end. It also should be remembered that the Professor was one of the panel before which Dr. Bush defended her thesis. It was apparent from his cross-examination that he would have strongly discouraged the publication of any PhD thesis which contained irrational speculation. None of the Bush thesis was criticised on this basis.

(c) Only a combined treatment would be thought of

150. On this issue, again the teaching of Bush appears to go further or be clearer than Rajfer. Her conclusion is that specific cGMP PDE inhibitors should be tried as therapeutic agents for the treatment of impotence. She makes the logical suggestion that papaverine, a known and widely used anti-impotence drug works in this way. It was well known that in many patients papaverine procured erections when applied alone. It seems to me that a skilled addressee would conclude that other cGMP PDE inhibitors might well function as papaverine does, either alone or in combination with other agents. This view is consistent with the immediately following suggestion that "a specific [cGMP PDE ] inhibitor could … produce erection…"

(d) and (e) No reason to find the Bell applications and Oral administration would be shunned

151. There is nothing in Bush which appears to take these issues any further than Rajfer.

(f) Others, including Pfizer, did not hit on the idea and commercial success

152. Once again, whatever may have been the teaching of Rajfer, Dr. Bush clearly had thought of the possibility of using specific cGMP PDE inhibitors in the treatment of impotence and said as much. The thesis was filed in the dissertation office of UCLA in December 1992 and a copy was also sent to the UCLA Bio Medical Library on 19 May 1993. The reality is that, although Pfizer accepts for the purpose of this action that the thesis was available to the public, dissemination is likely to have been quite restricted. In any event the period between publication and the priority date must be measured in weeks or months. For reasons already given, this undermines Mr Kitchin's argument of commercial success.

Conclusion in relation to Bush

153. Mr Thorley's submissions of obviousness succeed on this document as well.

Overview in relation to obviousness

154. The issue of obviousness can be looked at in another way. At paragraph 25 above is set out a passage in the patent which summarises the advance claimed by the inventor. It says that inhibitors of cGMP PDEs cause elevated cGMP levels and that the inventor has found that "unexpectedly … these compounds are useful in the treatment of erectile dysfunction". However in the light of each of Rajfer, Murray and Bush this is not an unexpected discovery. On the contrary it is just what each of those papers would have led a man in the art to suspect was the case. That suggestion is a central part of the express teaching of both the Murray article and the Bush thesis. The idea of using these inhibitors in the treatment of MED was made available to the scientific community by the ground-breaking work done by Professor Ignarro and his team. It was they who discovered the essential facts which put the design and production of convenient anti-impotence pills within the reach of the pharmaceutical world. It is not surprising that the publication of the Rajfer article produced the reaction it did (see paragraph 78 above). Even among members of the public it raised the promise of a pill for the treatment of impotence. It would have had the same effect on a man skilled in the art, save that he would not have thought in terms of a 'nitric oxide pill' alone. In the case of the Murray article and the Bush thesis he would have been pointed directly towards a PDE inhibitor pill.

ANTICIPATION

155. In the light of the findings made in respect of Rajfer, Murray and Bush, it is not necessary to consider the argument of anticipation based on Korenman at great length. Much effort and experimentation was put into this part of the case and it developed a size and complexity out of proportion to it importance in the case. In view of the likelihood of this case going further, I can set out my conclusions briefly.
156. Korenman's stated objective was to evaluate the use of a chemical called pentoxifylline to treat impotence in men with mild to moderate vascular insufficiency. Lilly says that pentoxifylline is a cGMP PDE inhibitor and that Korenman is therefore describing the administration of such an inhibitor to males for the purpose of treating their erectile dysfunction. Pentoxifylline is a well known compound although it has rarely been referred to as a PDE inhibitor. Pfizer says that this is not an accident. It argues that, at least at the doses used in Korenman, pentoxifylline is not a PDE inhibitor and it is certainly not potent or selective. Experiments were performed by the parties either to demonstrate efficacy or lack of it. What was quite apparent was that if pentoxifylline was acting as an inhibitor, it was not a pronounced effect. Pfizer adduced evidence from an expert in medical statistics, Professor Macrae, to the effect that the experiments conducted by Lilly did not show any statistically significant effect of pentoxifylline. Lilly did not call its own statistician. Instead it concentrated on subjecting Professor Macrae to detailed cross-examination. I found the Professor a fair and convincing witness, but it is not necessary to go into the detail of most of the issues touched upon in his evidence. Instead I can concentrate on one point.
157. It will be recalled that I have come to the conclusion that the claims of the patent require the PDE inhibitor to be effective in treating male erectile dysfunction. If a piece of prior art fails to demonstrate such efficacy, it cannot anticipate. Korenman reports experiments carried out on 24 couples where the male partner suffered from MED. 6 of the couples failed to complete the trials, so the results such conclusions as the authors drew were based on the sexual performance of the remaining 18 men. These were given pentoxifylline orally. Their sexual performance taking pentoxifylline was compared with performance when taking a placebo. Unfortunately 6 of this group made no attempts as intercourse on placebo or pentoxifylline despite the expressed interest of both partners in resuming coitus. Various reasons for this were given. Therefore such results as were obtained were based on the performance of 12 men. Dr Ellis gave evidence in relation to this. He said that other workers in the field had tried pentoxifylline for the treatment of MED. He referred in particular to papers by Georgitis & Merenich in 1995 and Knoll et al in 1996. He said that, viewed properly, Korenman does not claim that pentoxifylline works to treat MED. All it says is that the results justify doing other research to see if it works. This is the effect of the final paragraph in the paper:

"The small size of the study was partly due to the high proportion of dropouts and to the difficulty of continuing to accrue couples due to a logistical problem as a result of the authors' move to another institution. These promising results and the absence of side effects support additional clinical trials of pentoxifylline for this purpose."

158. The views of Georgitis, Merenich and Knoll do not prove that pentoxifylline does not work to cure MED. They demonstrate the beliefs of some workers in the field. What I am concerned with is whether pentoxifylline does discernibly cure MED, whatever the state of workers' beliefs. The onus of proving this is on Lilly and I only have to consider that issue on the basis of the material put before the court.
159. As I have said, the dispute between the parties on this issue has been extensive and complex. Lilly have conducted experiments to demonstrate that pentoxifylline is a selective cGMP PDE_V inhibitor. This has resulted in a profound clash on the evidence. However there are some points which can be made without getting too far into the detail of this part of the case. First, if pentoxifylline were working as a PDE inhibitor in the Korenman paper, it would be expected that erections would almost immediately be experienced in the successfully treated males. NO released on sexual stimulation would generate cGMP and, because the inhibitor would be effective to prevent the destruction of that chemical, erections should be produced or enhanced with little delay. However Korenman reports that there were weeks of delay before any erections were experienced. This produced a series of possible explanations from Lilly none of which was shown to be correct. Secondly, if pentoxifylline was acting as a cGMP PDE inhibitor, one would expect to see a rise in cGMP levels in corpus cavernosum tissue to which the chemical was applied. Such an experiment was carried out by Lilly but the repeat was abandoned. Therefore there is, on the material before me, no evidence that pentoxifylline is acting as a PDE inhibitor. Thirdly Pfizer advanced a number of criticisms of Korenman's experiments. I have already mentioned the small number of patients who took part in and completed the trials. Amongst the others, Pfizer point out that Korenman used a very low hurdle to indicate some form of anti-MED activity, namely one episode of sexual intercourse in three months. Further Korenman only relied on a subjective questionnaire to determine whether significant anti-MED action was demonstrated. In my view these points are telling. However they should not be read as a criticism of Korenman's paper. Read fairly Korenman does not assert that he had proved a discernible anti-MED effect of pentoxifylline. All that he said was that his results were such that it would be worthwhile to carry out proper trials. Such trials, if carried out, would no doubt include a much larger sample of subjects and would involve all the usual controls to minimize the chance of a spurious result. However, on the material presented in this trial, Lilly have not discharged the onus on them of proving that Korenman obtained a discernible anti-MED effect with pentoxifylline. It follows that the case of anticipation fails.
160. Notwithstanding this finding, Lilly succeeds in this action. The Pfizer patent is invalid for obviousness.

Note 1   In large part reproduced from ‘Penile Function’, Chapter 7 in ‘Pathophysiological principles of erection’, Blackwell Scientific, 1994.    [Back]