Teva Pharmaceutical Industries Ltd & Anor v Merrell Pharmaceuticals Inc & Anor

This case concerns an anti-histaminic drug called terfenadine and its acid metabolite, terfenadine carboxylate, which also has the generic name fexofenadine. Terfenadine features in litigation which found its way to the House of Lords: see Merrell Dow v Norton [1996] RPC 76. As Lord Hoffmann described the position in paragraph 1 of his speech:

“Merrell Dow Pharmaceuticals Inc. is a U.S. company with a United Kingdom subsidiary. I shall call them both "Merrell Dow". About 25 years ago Merrell Dow discovered an anti-histamine drug called terfenadine. It is used by people who suffer from hay fever and similar allergies and has the advantage that, unlike some other anti-histamines, it does not have the side-effect of making one drowsy. In 1972 Merrell Dow obtained a patent for terfenadine in the United Kingdom. After a period of extension under the Patents Act 1977, it finally expired in December 1992. Other pharmaceutical companies then started to make and market terfenadine.

In these proceedings Merrell Dow claims that their monopoly in terfenadine continues by virtue of a later patent which still has another 5 years to run. It was obtained in the following circumstances. After they had patented terfenadine, they did some research into the way it worked. They found that it passed through the stomach to be absorbed in the small intestine and was then 99.5% metabolised in the liver. This was why it had no side-effects. They analysed the chemical composition of the acid metabolite formed in the liver. Its chemical name is 4-[4(4-hydroxydiphenylmethyl-1-piperidinyl)-1- hydroxybutyl]-α, α-dimethylbenzene-acetic acid, but I shall call it the acid metabolite. No one had identified it before. So they patented the acid metabolite as claim 24 of a patent granted in 1980 for a number of related anti-histamine products. This is the patent in suit.”

There is no dispute about the correct approach to construction, novelty and obviousness.

In Kirin Amgen v TKT [2005] RPC 169, the House of Lords made it clear that the determination of the extent of protection conferred by a patent is an examination in which there is only one compulsory question, namely that set by Article 69 of the EPC and its Protocol: what would a person skilled in the art have understood the patentee to have used the language of the claim to mean?: see Lord Hoffmann at paragraph. 69.

There are two requirements for a claim to be anticipated by an earlier published document. First, the earlier document must disclose the invention claimed in the patent in suit. Secondly, the skilled reader must be able to perform the claimed invention if he attempts to do so by using the matter disclosed in the earlier document and/or his common general knowledge: see Synthon BV v SmithKline Beecham plc [2006] RPC 10. In Synthon, the well known judgment of the Court of Appeal in General Tire & Rubber Co v Firestone Tyre & Rubber Co Ltd [1972] RPC 457 at 485 concerning the issue of disclosure was approved by the House of Lords. See further the citation from Lord Hoffmann’s speech at paragraph 7-28 of Terrell on the Law of Patents (16^th ed)

A prior art document has to be construed as at the date it was published: see Jacob LJ in SKB v Apotex [2005] FSR 23 at para 89. This means that the document has to be read in the light of the common general knowledge as of its date, not as of the priority date of the patent in suit.

The structured (but not compulsory) approach to obviousness is set out in Windsurfing International v Tabur Marine [1985] RPC 59 at 73. It has recently been restated by Jacob LJ in Pozzoli SpA v BDMO SA[2007] EWCA Civ 588. It is worth setting out what he says there:

“Assessing obviousness

14 The place of "inventive concept" in relation to obviousness also calls for some discussion. It will be recalled that it forms the first step of the well-known Windsurfing test of Oliver LJ [1985] FSR 59 at 73. The test provides a structured approach to the problem and is often useful. I set it out adding my own numbering:

(1)

The first step is to identify the inventive concept embodied in the patent in suit.
(2) Thereafter, the court has to 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 that date, common general knowledge in the art in question.
(3) The third step is to identify what, if any, differences exist between the matter cited as being "known or used" and the alleged invention.
(4) Finally, the court has to ask itself whether, viewed without any 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.

15 I think the test requires some restatement and elaboration. First one must actually conduct the first two operations in the opposite order -- mantle first, then concept. For it is only through the eyes of the skilled man that one properly understand what such a man would understand the patentee to have meant and thereby set about identifying the concept.

16 Next, that first step actually involves two steps, identification of the attributes of the notional "person skilled in the art" (the statutory term) and second identification of the common general knowledge ("cgk") of such a person.

17 What now becomes stage (2), identifying the inventive concept, also needs some elaboration. As I pointed out in Unilever v Chefaro [1994] RPC 567 at page 580:

”It is the inventive concept of the claim in question which must be considered, not some generalised concept to be derived from the specification as a whole. Different claims can, and generally will, have different inventive concepts. The first stage of identification of the concept is likely to be a question of construction: what does the claim mean? It might be thought there is no second stage -- the concept is what the claim covers and that is that. But that is too wooden and not what courts, applying Windsurfing stage one, have done. It is too wooden because if one merely construes the claim one does not distinguish between portions which matter and portions which, although limitations on the ambit of the claim, do not. One is trying to identify the essence of the claim in this exercise.”

18 So what one is seeking to do is to strip out unnecessary verbiage, to do what Mummery LJ described as make a précis.

19 In some cases the parties cannot agree on what the concept is. If one is not careful such a disagreement can develop into an unnecessary satellite debate. In the end what matters is/are the difference(s) between what is claimed and the prior art. It is those differences which form the "step" to be considered at stage (4). So if a disagreement about the inventive concept of a claim starts getting too involved, the sensible way to proceed is to forget it and simply to work on the features of the claim.

20 In other cases, however, one need not get into finer points of construction -- even without them the concept is fairly apparent -- in Windsurfing, for instance, it was the "free sail" concept. In yet other cases it is not even practical to try to identify a concept -- a chemical class claim would often be a good example of this.

21 There is one other point to note. Identification of the concept is not the place where one takes into account the prior art. You are not at this point asking what was new. Of course the claim may identify that which was old (often by a pre-characterising clause) and what the patentee thinks is new (if there is characterising clause) but that does not matter at this point.

22 The third step also requires a little reformulation -- Windsurfing was a case under the 1949 Act where the statutory words for the prior art were "known or used". The European Patent Convention uses the words "state of the art".

23 The fourth step needs no restatement, though it is worth making explicit that by invention is meant what is claimed. In the result I would restate the Windsurfing questions thus:

(1)

(a) Identify the notional "person skilled in the art"

(b)

Identify the relevant common general knowledge of that

person;

(2)

Identify the inventive concept of the claim in question or if that cannot readily be done, construe it;

(3)

Identify what, if any, differences exist between the matter cited as forming part of the "state of the art" and the inventive concept of the claim or the claim as construed;

(4)

Viewed without any knowledge of the alleged invention as claimed, do those differences constitute steps which would have been obvious to the person skilled in the art or do they require any degree of invention?”

The Court must be particularly careful not to allow hindsight to colour its judgment when considering the fourth step. Oliver LJ himself stressed this in Windsurfing, as have many cases before and since.

The present case, too, raises the familiar “obvious to try” argument Mr Waugh refers, in this context, to the expectation the skilled man would have of the success of his postulated trials as a relevant consideration. He cites Saint-Gobain PAM SA v Fusion Provida Ltd [2005] EWCA Civ 177, where Jacob LJ said this about such arguments in paragraph 35 of his judgment:

“35.

None of this to my mind remotely makes the idea of using Zn/Al alloy for pipes obvious – as something which is simply self-evident to the unimaginative man skilled in the art. Mere possible inclusion of something within a research programme on the basis you will find out more and something might turn up is not enough. If it were otherwise there would be few inventions that were patentable. The only research which would be worthwhile (because of the prospect of protection) would be into areas totally devoid of prospect. The "obvious to try" test really only works where it is more-or-less self-evident that what is being tested ought to work”.

As I said in Actavis v Merck, [2007] EWHC 1311 (Pat) it is often dangerous to take short passages from a judgment out of context, and what Jacob LJ says here really needs to be read in the context of the facts of the case. In any event, at the end of the day, the “obvious to try” argument is directed at, and is only one component in answering, the single question whether the invention was obvious. As Jacob LJ himself put it, after citing St Gobain, in his judgment in Angiotech Pharmaceuticals v Conor Medsystems Inc [2007] EWCA Civ 5, at paragraph 45:

“In the end the question is simply "was the invention obvious?" This involves taking into account a number of factors, for instance the attributes and cgk of the skilled man, the difference between what is claimed and the prior art, whether there is a motive provided or hinted by the prior art and so on. Some factors are more important than others. Sometimes commercial success can demonstrate that an idea was a good one. In others "obvious to try" may come into the assessment. But such a formula cannot itself necessarily provide the answer. Of particular importance is of course the nature of the invention itself.”

It is worth noting that in Angiotech, Jacob LJ goes back to the source of the expression “obvious to try” in relation to obviousness which is found in the judgment of Diplock LJ in the Johns-Manville case, [1967] RPC 479. Jacob LJ says this:

“The facts were simple: there was a known process. The patent was for the old process using the new agent. It was held obvious as being “well worth trying out”. Diplock LJ said:

‘It is enough that the person versed in the art would assess the likelihood of success as sufficient to warrant actual trial.’ ”

There is, of course, a difference of emphasis, at least, between “well worth trying out” and “ought to work”. But in the context of the ultimate question, Was the invention obvious?, the difference may come to nothing. I do not think that Jacob LJ can be read as saying that they are inconsistent.

The fact that there may be several obvious courses of action does not prevent each of them being obvious: Brugger v Medic-Aid [1996] RPC 635 at 661 and see Terrell at 7-71.

Although obviousness is concerned with technical obviousness and not commercial obviousness, the issue nonetheless needs to be addressed in the real world, a world where commercial considerations are of great importance. The mindset of the skilled man will be influenced by these matters; and the mindset of the skilled man must be factored into the question of obviousness: see for instance Dyson v Hoover [2002] RPC 22 at paras 56 - 67 (Aldous LJ), 87-88 (Sedley LJ) and 93 - 97 (Arden LJ).

It will be seen that most of the Claims are in “Swiss form” (for a general description of which see Terrell at 6-122ff). I have recently addressed this topic in my judgment in Actavis v Merck and do not propose to repeat everything which I said there. I should, however, for completeness set out the relevant statutory provisions and address one authority.

Sections 4(2) and 4(3) of the Patents Act 1977, which are derived from Article 52(4) of the Convention on the Grant of European Patents (the European Patent Convention) (“EPC”), state:

“(2)

An invention of a method of treatment of the human or animal body by surgery or therapy or of diagnosis practised on the human or animal body shall not be taken to be capable of industrial application.

(3)

Subsection 2 above shall not prevent a product consisting of a substance or composition being treated as capable of industrial application merely because it is invented for use in any such method.”

Section 2(6) of the 1977 Act, which is derived from Article 54(5) of the EPC, states:

“(6)

In the case of an invention consisting of a substance or composition for use in a method of treatment of the human or animal body by surgery or therapy or of diagnosis practised on the human or animal body, the fact that the substance or composition forms part of the state of the art shall not prevent the invention from being taken to be new if the use of the substance or composition in any such method does not form part of the state of the art.”

The result is that it is possible to obtain a patent for the first medical use of a known substance or composition, where this substance or composition was not previously known to have any medical application: see Sopharma SA’s Application [1983] RPC 195.

Owing to the prohibition on method of treatment claims in Article 52(4) of the EPC (which corresponds to sections 4(2) and 4(3) of the 1977 Act), the Enlarged Board of the European Patent Office has held that a European patent may notbe granted for the use of a substance or composition for the treatment of the human or animal body by therapy.

However, it has also accepted, on policy grounds, and following the practice of the Swiss Federal Intellectual Property Office, that a European patent may be granted with claims directed to the use of a substance or composition for the manufacture of a medicament for a specified new and inventive therapeutic application. This gave rise to the now widely used form of the second medical use, or so-called “Swiss form” claim: see Eisai/Second Medical Indication (1985) OJEPO, 64 (“Eisai”).

Accordingly, a claim in the form: “Use of [X] for treatment of [Y]” would not be accepted, whereas a claim “Use of [X] for the manufacture of a medicament for treatment of [Y]” would be accepted. It might be noted that the feature supporting novelty and inventive step is the new treatment; nonetheless the Swiss form claim itself covers the use in preparation of the medicament in question, not the use of the medicament for the novel therapy.

The leading case under English law in relation to claims in the Swiss form is the decision of the Court of Appeal in Bristol-Myers Squibb v Baker Norman [2001] RPC 1. This case concerned a novel dosing regime rather than a new therapeutic application. I referred to this case and the reasoning of the judges as set out at paragraphs 34-49 and paragraphs 54-63 (Aldous LJ), paragraphs 76-88 and 90-94 (Buxton LJ) and paragraphs 107-112 (Holman J) in my judgment in Actavis v Merck.

In summary, the Court of Appeal held that, in so far as Swiss-type claims were permissible at all, they must be limited to a therapeutic application which was not only inventive but new; novelty must reside in the new second, or subsequent, therapeutic use. In other words, the novelty must lie in a new application (ie the known compound is to treat an additional disorder or prevent a disease rather than cure) not in an improved use in the method of administering an existing treatment.

Mr Waugh points out, however, that the Technical Board of Appeal in the EPO has recently interpreted the Eisai decision differently to the Court of Appeal in Bristol-Myers and expressly declined to follow it. He says that the law in this country is now out of step with the EPO. In the light of the observations of the House of Lords to the effect that even decisions of the EPO, and a fortiori of the Technical Board of Appeals, are to be regarded as of great persuasive authority, it is said that I should not follow Bristol-Myers.

I do not consider that it is open to me to depart from the decision in Bristol-Myers which is a carefully considered and argued decision of the Court of Appeal. It is for that Court to consider whether its earlier decision should be departed from in the light of developing case-law in the EPO.

It is common ground that the skilled addressee is a team working in a pharmaceutical company. Thus Professor Frew identifies the team as comprising medicinal and formulation chemists, clinicians and pharmacologists. Mr Waugh identifies a more specific list: his team would comprise a chemist familiar with the chemistry of histamine (H₁) receptor antagonists, biologists with practical experience of the techniques used to characterise receptor agonists and with experience of animal models to measure inter alia skin reactions to histamine and allergens. The team would also have co-opted clinical investigators, toxicologists and pathologists according to need. The team would also include a biologist familiar with the electrophysiological recording techniques.

In order to understand the patents in suit and some of the prior art cited and other articles (some of which are part of the cgk) I need to explain what is meant by the QT interval and Ik channel. Professor Frew gives a helpful explanation which I adopt as follows.

In his first report, Professor Frew includes at Annex 12 an explanation of ventricular rhythm disturbances which is not a matter of controversy. He explains that cardiac muscles have a resting potential of -60 to -90 millivolts, due to a difference in the ionic content of the cell and the extracellular fluid. Upon activation, a series of short lived changes occur in the permeability of the cell membrane to sodium, potassium and calcium ions. He explains that depolarisation initially increases the membrane permeability to sodium which flows into the cell passively (sodium concentrations in the cell being lower than outside). Depolarisation is terminated by closure of the sodium channels and opening of a set of voltage-dependent potassium channels which allow potassium to exit the cell (potassium concentrations being much higher inside the cell than outside). Next there is a plateau phase in which calcium ions enter the cell for a variable period – between 150-400 milliseconds. The final phase is re-polarisation which occurs when the calcium channels close and the outward potassium current increases. This is the delayed rectifier potassium current Ik. This potassium flux restores the normal resting potential across the membrane until the next depolarising impulse arrives. During the resting phases ionic pumps work to exchange sodium and potassium ions to restore normal balance before the next depolarisation wave arrives.

Professor Frew includes in the Annex an ECG in which depolarisation is seen as the “QRS complex” while repolarisation forms the “T wave”. The QT interval is the time between the start of the QRS complex and the end of the T wave. The duration of the depolarisation varies according to the heart rate, so that the QT interval has to be corrected for heart rate resulting in the “corrected QT interval” or QTc interval.

I start with 976. It has a priority date of 11 May 1992 and is entitled “Use of Terfenadine derivatives as antihistaminics in a hepatically impaired patient”. From the brief introduction which I have already given, it will have been noted that administering the acid metabolite addresses the risk of TdP associated with terfenadine itself in certain patients, in particular hepatically impaired patients. 976 is an attempt to achieve a monopoly in respect of that patient group by use of a Swiss form claim in order to circumvent any perceived concern about the non-patentability of second medical use claims.

976 starts by recognising the 1992 PDR and Carr. It refers to the fact that terfenadine undergoes extensive (99%) first pass metabolism, to an active acid metabolite and an inactive dealkylated one. It states that the acid metabolite has been disclosed by Carr as an antihistaminic agent having oral activity and notes that studies investigating the effect of hepatic and renal insufficiency on the metabolism and excretion of terfenadine are incomplete.

It reports that preliminary information indicates that in patients with hepatic impairment, significant concentrations of unchanged terfenadine can be detected with the rate of formation of the acid metabolite being decreased. In subjects with normal hepatic function, unchanged terfenadine plasma concentrations have not been detected. It also notes that it has recently been found that patients with hepatic impairment (alcohol, cirrhosis, hepatitis) or on ketoconazole or troleandomycin therapy, or having conditions leading to QT prolongation (eg hypokalemia, congenital QT syndrome) may experience cardiac events of QT prolongation and/or ventricular tachycardia at the recommended doses of terfenadine.

The Patent then states that “surprisingly” it appears that patients with impaired hepatic function who are receiving the acid metabolite in sufficient amount so as to provide an antihistaminic effect will not experience cardiac events of QT prolongation and/or ventricular tachycardia.

The Patent discloses that patients receiving the acid metabolite at a level sufficient to produce an anti-histaminic effect will not experience those cardiac side effects seen in hepatically impaired patients receiving terfenadine. In this context, a hepatically impaired patient is one having impaired liver function due to diseases or due to administration of a drug such as ketoconazole which inhibits normal liver metabolic function.

The Patent then provides under the heading “Summary of the Invention” that the present invention relates to the use of a compound of Formula (1) which is then described. Coming to material under the heading “Detailed Description of the Invention”, the Patent states that compounds of Formula (1) (and in particular, a specified formulation of it) are prepared and used as described in Carr: see paragraph [0008].

Paragraph [0009] tells us that the invention “relates to the use of a compound of formula (1) in the manufacture of medicaments for providing an antihistaminic effect in a hepatically impaired patient in need thereof comprising administering to said patient an effective antihistaminic amount”. And paragraph [0010]goes on to explain that such compounds are known histamine H₁-receptor antagonists and as such provide relief of symptoms associated with histamine-mediated diseases and conditions such as seasonal allergic rhinitis, urticaria and the like.

Paragraph [0013] is a useful description of what is said to be surprising:

“When administered terfenadine at the recommended dosages, a hepatically impaired patient will experience increased levels of terfenadine in the blood and decreased levels of the acid metabolite over that expected with the non-hepatically impaired patient. Increased blood levels of terfenadine may in turn cause decreases in the action potential and in various membrane currents of cardiac cells which may trigger cardiac events of QT prolongations and/or ventricular tachycardia. Surprisingly, similar blood levels of the terfenadine acid metabolite do not cause these decreases in the action potential and in various membrane currents of cardiac events of QT prolongations and/or ventricular tachycardia, the terfenadine acid metabolite will not trigger these cardiac events.”

The relevant claims of 976 are:

“1.

The use of a compoundof the formula [depicted and defined being a class within which the acid metabolite falls] or a pharmaceutically acceptable salt and individual isomers thereof in the manufacture of a medicament for providing an anti-histaminic effect in a hepatically impaired patient in need thereof.

5.

The use of claim 1 wherein the compound is [the acid metabolite].

8.

The use of a compound of the formula [depicted and defined, being a class within which the acid metabolite falls] or a pharmaceutically acceptable salt and individual isomers thereof in the manufacture of a medicament for providing an anti-histaminic effect in a patient in need thereof while avoiding the cardiac events associated with the administration of terfenadine to a hepatically impaired patient.

9.

The use of claim 8 wherein the compound is [the acid metabolite].

12.

Use according to any one of claims 1-11 for the preparation of a pharmaceutical composition for the treatment of seasonal allergic rhinitis and urticaria.”

I come next to 443. It has a priority date of 3 August 1992 and is entitled “Terfenadine metabolites and their optically pure isomers for treating allergic diseases”. It relates to compositions containing the acid metabolite.

The Patent refers (in paragraph [0010]) to the metabolism of terfenadine: it is well absorbed but is extensively metabolised (reference being made to a 1981 paper by Okerholm and others). It notes the suggestion that the acid metabolite may show in vitro anti-histaminic activity but that no data had been published. In paragraphs [0015] and [0016] reference is made to terfenadine causing less of the usual side effects of anti-histamines.

Paragraph [0017] to [0019] are of some significance and I set out the material parts:

“[0017] However, the administration of terfenadine to a human has been found to cause other adverse effects. These adverse effects include but are not limited to cardiac arrhythmias, including ventricular tachyarrhythmias, torsades de pointes, and ventricular fibrillation. Recently, clinical practitioners have noted an increase in the occurrence of these cardiac arrhythmias upon coadministration of terfenadine with other drugs such as ketoconazole and erythromycin or upon overdoses of terfenadine. See [Monahan] and Sandra Knowles [citation given].

[0018] Thus, it would be particularly desirable to find a compound with the advantages of terfenadine which would not have the aforementioned disadvantages.

[0019] It has now been discovered that [ compounds including the acid metabolite]….are useful for treating allergic rhinitis without inducing the cardiac arrhythmias associated with terfenadine.”

Example 3 at paragraph [0052] provides data comparing the propensity of terfenadine to block delayed rectifier potassium current with absence of such an effect by the acid metabolite (the tests being carried on ventricular myocytes obtained from isolated cats hearts by conventional techniques). The conclusion drawn (in paragraph [0053] ) is the acid metabolite “surprisingly” is not [this might perhaps better read “may not be”]liable to cause cardiac arrhythmia at dose levels at which there is a distinct risk of such a side effect being caused by terfenadine itself.

Paragraph [0022] tells us that prior to the present invention, those skilled in the art would have expected the acid metabolite to have induced TdP (as reported in Monahan and Sandra Knowles). This is so, according to the Patent, because the potentially lethal arrhythmia was considered to be a “class effect” among non-sedating anti-histamines, in the sense that the arrhythmogenicity was considered to be coupled to the anti-histaminic potency of such compounds: there is a dispute about that. It goes on to say this:

“Accordingly, the fact that the compositions, in accordance with the present invention, do not induce any such cardiac arrhythmias is a new, highly useful and surprising technical effect, which enables the inventive compositions to be administered to individuals susceptible to cardiac arrhythmia, and in potentially larger doses that those non-sedating anti-histamines, such as terfenadine, in common use at the present time.”

The relevant claims of 443 are:

“1.

Use of a composition comprising [the acid metabolite] or a pharmaceutically acceptable salt thereof for the preparation of a medicament for use in a treatment of allergic rhinitis in which the induction of cardiac arrhythmia is avoided, said treatment comprising administering a therapeutically effective amount of [the acid metabolite] to a human patient whose hepatic function is not impaired.

12.

A pharmaceutical composition in the form of an oral solid preparation comprising a unit dosage of 60mg of [the acid metabolite] or a pharmaceutically acceptable salt thereof, and a pharmaceutically acceptable carrier or excipient, for use in a treatment of allergic rhinitis in which the induction of cardiac arrhythmia is avoided.”

The final Patent is 860. It has a priority date of 3 August 1992 and is entitled “Terfenadine carboxylate and the treatment of allergic disorders”. It is a divisional application from 443. It therefore has essentially the same disclosure as 443.

The relevant claims are:

“1.

Use of a composition comprising [the acid metabolite] or a pharmaceutically acceptable salt thereof, for the manufacture of a medicament for use in an anti-histaminic treatment in which the induction of cardiac arrhythmia is avoided, said treatment comprising administering a therapeutically effective amount of [the acid metabolite] to a human patient whose hepatic function is not impaired.

3.

A use as claimed in claim 2, wherein the allergic disorder is asthma or allergic rhinitis.”

Each of the Patents share one aspect of the inventive concept, namely the use of the acid metabolite to manufacture a medicament for the treatment of the identified conditions and thereby avoid the cardiac side effects which had occurred, according to Mr Waugh, in both hepatically impaired and unimpaired patients.

Teva’s attack before me on the validity of the Patents is based on the acid metabolite. It is accepted by the Defendants that if claim 5 of 976 is invalid, so too is claim 1; likewise with claims 9 and 8. It is also accepted that the validity of claims 1 and 3 of 860 falls with claim 1 of 443.

I agree with Mr Watson that the technical content of the patents is at a preliminary level:

Monahan. This appeared in December 1990 but it reports an event from a year earlier. It was a case of a woman who was undergoing terfenadine therapy and who developed a prolonged QT interval and TdP shortly after starting concomitant therapy with a drug called ketoconazole as indicated by the figures on the second page of the report. Ketoconazole was a well-known inhibitor of liver metabolism. The abstract of Monahan states:

“….Although found in many clinical settings, torsades de pointes is most often drug induced. This report describes the first association (exclusive of drug overdose) of symptomatic torsades de pointes occurring with the use of terfenadine in a patient who was taking the recommended prescribed doses of this drug in addition to ceflacor, ketoconazole, and medroxyprogesterone. Measured serum concentrations of terfenadine and its main metabolite showed excessive levels of parent terfenadine and proportionately reduced concentrations of metabolite, suggesting inhibition of terfenadine metabolism. We believe that a drug interaction between terfenadine and ketoconazole resulted in the elevated terfenadine levels in plasma and in the cardiotoxicity previously seen only in cases of terfenadine overdose.”

On the first page of the report, after noting that cardiotoxicity from terfenadine has only been reported in overdose, the authors continue:

“This report describes the first association in a patient who was taking the recommended does of terfenadine. We believe a drug interaction between terfenadine and ketoconazole was responsible for this striking cardiac dysrhythmia.”

Then, in the comment section later in the report, the authors say this:

“Thus far, the literature is devoid of documented serious adverse effects or significant drug interactions that involve the H₁ antagonist, terfenadine. Terfenadine is highly metabolized by the liver, presumably by the P-450 oxidative pathways. Normally, serum concentrations are below assay detection limits (10 ng/ml) in patients who are taking a dosage of 60mg twice a day. We believe the serum concentrations of the parent drug and the main metabolite are the result of the inhibition of terfenadine metabolism. The marked increase in the serum concentration of the parent compound may have produced the arrhythmia seen in this case. Identical cardiotoxicity has been reported with acute terfenadine overdosage. The presence of a relatively low serum concentration of metabolite supports this hypothesis.”

Then, after warning of the dangers of this perceived drug interaction, they conclude by saying that further studies are indicated to elucidate the mechanisms of this adverse drug interaction.

The expert reports revealed this difference: Professor Frew, in his review of Monahan, explains why in his view it is made clear that it is unmetabolised terfenadine which is responsible for the cardiac side effects and that accordingly it was obvious to develop the acid metabolite. Dr Morley considers Monahan to be making unjustified assumptions; in any case, his paper was only one among many. He does not consider that the skilled person would consider investigating the properties of the acid metabolite. I will need to resolve this difference when reviewing their evidence.

The 1992 PDR. The PDR is an annual publication for US physicians, intended to provide a summary for non-specialists of what is known about each drug. It is common ground that this is material that the skilled team would know of.

Under “Clinical Pharmacology” one finds the following:

“Terfenadine itself undergoes extensive (99%) first pass metabolism to two primary metabolites, an active acid metabolite and an inactive dealkylated metabolite. Therefore, systemic availability of terfenadine would be low.

…

Preliminary information indicates that in cases of hepatic impairment, significant concentrations of unchanged terfenadine can be detected with the rate of acid metabolite formation being decreased. In subjects with normal hepatic function unchanged terfenadine plasma concentrations have not been detected.

In vitro studies demonstrate that terfenadine is extensively (97%) bound to human serum protein while the acid metabolite is approximately 70% bound to human serum protein. Based on data gathered from in vitro models of antihistaminic activity, the acid metabolite of terfenadine has approximately 30% of the H₁ blocking activity of terfenadine. The relative contribution of terfenadine and the acid metabolite to the pharmacodynamic effects have not been clearly defined. Since unchanged terfenadine is usually not detected in plasma and active acid metabolite concentrations are relatively high, the acid metabolite may be the entity responsible for the majority of efficacy after oral administration of terfenadine.”

The section in the 1992 PDR on Precautions contains the following:

“General

Terfenadine undergoes extensive metabolism in the liver. Patients with impaired hepatic function (alcoholic cirrhosis, hepatitis) or on ketoconazole or troleandomycin therapy, or having conditions leading to QT prolongation (e.g. hypokalemia, congenital QT syndrome) may experience QT prolongation and/or ventricular tachycardia at the recommended dose. The effect of terfenadine in patients who are receiving agents which alter the QT interval is not known. These events have also occurred in patients on macrolide antibiotics, including erythromycin, but causality is unclear. The events may be related to altered metabolism, to electrolyte imbalance, or both.

…

Drug interactions

Preliminary evidence exists that concurrent ketoconazole or macrolide administration significantly alters the metabolism of terfenadine. Concurrent use of Seldane with ketoconazole or troleandomycin is not recommended. Concurrent use of other macrolides should be approached with caution.

Under “Adverse Reactions”, reference is made to rare reports of severe cardiovascular adverse effects including arrhythmias including TdP.

Chen. Chen reports the effect of terfenadine on the rectifier potassium current I_k using the cat ventricular myocytes test (the test used in Example 3 of 443 / 860). The conclusion is as follows:

“Terfenadine is a potent blocker of I_k. This action may be responsible for the rare cases of QT interval prolongation observed at high dosages and torsades de pointes seen with drug interactions that produce high levels of terfenadine.”

Carr. Carr discloses a number of compounds which, it also discloses, are useful as antihistamines, anti-allergy agents and bronchodilators as well as methods of making and using the same. One of the compounds is the acid metabolite. This is the subject of example 3 (how to make the acid metabolite) and of example 12 (an illustrative composition for an aerosol suspension). The compound is claimed in claim 8.

The acid metabolite is also the compound referred to in order to demonstrate the utility of the compounds of the invention. It is said that the acid metabolite gives a significant reduction in histamine induced isolated guinea pig ileal muscle contraction (referred to elsewhere in the present case as the guinea pig ileum test, a standard assay for antihistamine activity).

I would only add at this point that in 1992 it was common general knowledge that the acid metabolite was produced on administration of terfenadine and was, at the least, believed to make a significant contribution to its efficacy as an anti-histamine.

The 976 Application. This is not now separately relied on.

The first event which I need to record is a case report from Dr Craft, senior house office at a hospital in Portsmouth England. This reports a case of acute ventricular arrhythmia that occurred after astemizole overdose. Astemizole is another second generation antihistamine. The patient recovered and was discharged with advice not to take astemizole. Follow up in outpatient department confirmed that she was well with a normal QT interval.

There were subsequent reports of arrhythmia including TdP in 1988 and 1989 which were all astemizole overdose cases. Then, in a letter from Dr Davies and others, British Medical Journal 1989 298: 325 (headed Cardiotoxic effect with convulsions in terfenadine overdose) a similar problem was reported in what appears to be a case of terfenadine overdose with high plasma levels of the acid metabolite and unmetabolised terfenadine. The raised concentrations coincided with prolonged QT interval.

Although Monahan was not published until December 1990, the case which it reports had occurred in November 1989 and was known to Merrell Dow. Whether or not as a result of the findings of prolongation of QT interval with terfenadine, and in particular the Monahan case, an FDA hearing was called by the Pulmonary - Allergy Drugs Advisory Committee on 11 June 1990.

Merrell had made available to the Committee an internal paper entitled “Review of the Cardiovascular Effects of Terfenadine”. That paper contains an introduction which gives a brief history of the marketing of terfenadine and tells the reader that, as of May 1990, Merrell Dow had received some 3,000 postmarketing reports of adverse drugs experience. These included 6 cases of QT interval prolongation, 12 cases of QT interval prolongation accompanied by TdP and 7 cases of other types of cardiovascular events. 8 of these 25 cases occurred with overdose; 17 occurred at normal dosages but were in patients with pre-existing disease states or who were taking concomitant medications known to be associated with ventricular arrhythmias. There follows a detailed review of the cases.

Of particular relevance are tables 1 and 3. Table 1 contains a summary of 8 reported postmarketing overdose cases with prolonged QT and/or ventricular arrhythmias. Table 3 contains a summary of 6 cases of cardiovascular adverse effects in patients with CV risk factors associated with alcoholism and/or cirrhosis.

As to table 1, there are 3 cases where it is recorded “no parent compound detected” and 1 where it is recorded “no parent compound”. The latter was a case of TdP in the case of a patient with alcoholism and cirrhosis with ascites. The former 3 cases recorded no concomitant disease. They recorded QT prolongation but no TdP. Table 3 records a single case where it is recorded “no parent compound detected”. The adverse events in relation to this patient were ventricular fibrillation, ventricular tachycardia with TdP and prolonged QT.

The matter brought to the committee was terfenadine’s ability to cause prolongation of the QT interval and TdP. It appeared from the opening remarks of Dr Botstein that the

“same problem may well exist with other second-generation antihistamines. Astemizole… a recently marketed second-generation antihistamine, can also cause torsades de pointes under some circumstances. We know of a third second-generation antihistamine which is in early clinical trials and appears to cause, under some circumstances, prolongation of the QT interval.”

There followed a presentation by Dr Rossi. It is unfortunately not always possible to tell precisely what message he is conveying because his presentation was accompanied by transparencies which have not been made available. It is not possible or sensible to begin to summarise the presentation in this judgment. But a number of points which come out of it ought to be made:

There then followed a presentation by Dr Griffin of Merrell Dow. Again, I cannot summarise the presentation, but the following emerges:

In the discussion after the presentation the following, among other matters, emerge. First, Dr Griffin (of Merrell Dow) said that “We certainly believe that the majority, if not all, of the activity that we are seeing in humans, as far as antihistaminic activity is concerned, is coming from the acid metabolite. We are simply unable, except in rare circumstances to pick it up”. Standing back Dr Morley would say that there was evidence that the majority (but not all) of such activity was coming from the acid metabolite.

Next, Dr Lazzara (the cardiologist co-opted onto the panel) said he wanted to get clear whether it is the acid metabolite that might be implicated in the problem or whether it is the drug. In the overdose cases, he understood the acid metabolite to be greatly in excess and the parent compound is not detectable. A debate ensured. Dr Burke focused that debate when he asked “Are we dealing with a dose-related phenomenon? Is it related to the terfenadine? Is it related to the acid metabolite? Are other more important interactive factors operant here?” Dr Lazzara expressed his view in relation to that:

“I would think the clearest evidence is for the acid metabolite. The overdoses and the 300-mg dosing that showed the distinctly longer QTs – both would be predicted to have higher acid metabolite level, not necessarily much of a parent compound level. “

Dr Burke’s response to that was that many of the patients who had the overdose did not have the abnormality in the QT or the arrhythmia – it was “just a small subset of the overdose patients who are winding up with the cardiac abnormalities”.

Dr Lipicky said this:

“Whether it is the parent drug or the acid metabolite that is principally responsible for the QT interval prolongation I guess would depend on having data that had serial measurements of QT, serial measurements of terfenadine and the acid metabolite and then which one correlated better that the other. That information is totally lacking.”

Professor Woosley expressed a view contrary to that of Dr Lazzara, and had the impression that

“it more likely to be terfenadine that is the agent causing QT prolongation, based on some of the overdoses having measurable levels of the parent compound and some of the drug interaction cases have measurable, very high levels of the parent compound associated with QT prolongation.”

Later in the discussion, Professor Woosley raised the question of a class effect, saying that this particular drug might only be a representative of a class of drug. Dr Gothe agreed saying that she did not think this particular drug has to take all the blame. She also said that she was not convinced that terfenadine was that much different from other antihistamines.

Following the hearing, there appeared a report in Health News Daily on 13 June 1990. This document appeared in a bundle prepared by the Defendants’ solicitors for cross-examination of Professor Frew where it is indexed as Merrell Dow’s “Dear Doctor letter on cardiovascular side effects”. It is not, in fact, a letter at all. It started

“Marion Merrell Dow’s Seldane “Dear Doctor” letter on cardiovascular side effects should accompany revised labelling for the non-sedating antihistamine, FDA’s Pulmonary Drugs Advisory Committee recommended at its June 11 meeting.”

It contained the following statement:

“The company believes that the adverse events seen in the ketoconazole and cirrhosis patients were in effect overdose situations where the antibiotics/cirrhosis caused the patients to metabolise Seldane differently.…”

Although not apparently addressed to “Dear Doctor”, Merrell Dow subsequently sent a letter dated 6 August 1990 to US physicians enclosing a revised Product Information sheet for Seldane, dated July 1990, which together acknowledged the cardiac side effects. This document contains on the first full page of text the words "important drug warning" in a black-lined box. The document appeared as Annex 14 to Professor Frew’s first report. It also appears in a bundle prepared by the Defendants’ solicitors for cross-examination of Professor Frew. In that edition, there is an extra page which appears before the text of the letter – although it is not clear whether this in fact forms part of the same document and if so how it fits in – which also contains those words in a heavy black box.

The letter advised of a re-labelling and states that, although the events noted are extremely rare, it was considered that the changes were of sufficient importance to inform practitioners. The letter contains an interesting side-swipe at competitors:

“The FDA has suggested that this letter also inform you that the rare cardiovascular events reported in patients taking Seldane (QT prolongations and torsades de pointes) have also been reported in patients taking first and other second generation antihistamine drugs, particularly in overdoses.”

I will return to that statement in the context of the alleged class effect: see paragraphs 131ff below.

Under “Precautions” appears the following:

Following the drug warning, the Editor of Ear, Nose & Throat Journal arranged a teleconference to provide readers with answers to some of the questions which this threw up. This is recorded as Special Drug Communication Teleconference August 7 1992, inENT Journal September 1992, Vol 71 No 9 pages 383-390. One of the participants was Dr McNutt, project director for Seldane at Merrell Dow.

Dr McNutt refers to a “Dear Doctor” letter in August 1990, presumably the letter dated 6 August 1990 which I have just referred to. He gave a thumbnail sketch of reported instances of QT prolongation alone or associated with serious ventricular arrhythmias. A report from the Bethesda Naval Hospital in 1989 [the Monahan case] was the first time someone had taken nizoral and seldane together without any other underlying illness: the patient went into TdP. He is recorded as saying this:

“Really the first bellwether case was in 1989, reported by the Bethesda Naval Hospital ….We reviewed further and found that there were three other cases with ketoconazole. This led us to believe that perhaps there was some disruption going on with ketoconazole. Initially we didn’t think it was a disruption of the metabolism because we didn’t feel it hit exactly where Seldane was metabolized, i.e. in the cytochrome P-450 system in the liver. Still we went ahead and did a study (as we continued to accumulate other cases) and the outcome was that, when you gave a single dose of Seldane to someone who was already at steady state levels of ketoconazole, you had about a 30 msec increase in the QTc. We also saw an extraordinarily increased peak of the parent compound, which ordinarily isn’t present in the plasma in detectable levels.”

The next point of significance so far at least as concerns public perceptions is Monahan (in contrast with the case with which it deals which was known to the FDA at the 11 June 1990 committee meeting) which I have dealt with under the heading Prior Art, followed by a number of articles which I deal with later. [MacConnell and Stanners (15 June 1991), Zamani (accepted July 1991), Hoppu (31 August 1991), Mathews (6 November 1991), Knowles (February 1992)].

To complete the chronology, in May 1992 was the first filing in the US of 976 and in August 1992 filings in the US for 443 and 860. Then followed a number of papers which I also deal with later [Honig (September 1992), Woosley I (March 1993), Woosley II (1996), DuBuske (1996), Barbey (May-June 1999)]

Since it has featured in the argument, I need to say something about what it was that led to the FDA committee hearing in June 1990. Professor Frew’s understanding was that it was the Monahan case that precipitated the hearing. There is some evidential support for that.

First, there is a chronology produced by Merrell Dow in May 1990 of events involving terfenadine and the development of cardiac adverse events. The report of the Monahan event came to Merrell Dow on 11 November 1989. There had been some reports of adverse events surrounding terfenadine before 1989, but these did not give rise to undue concerns. It is stated in the chronology that, because of the case reported by Monahan, the relationship of ketoconazole was reviewed leading to discovery of other reports.

There had been an advisory committee meeting fixed for 7 December 1989 to discuss whether terfenadine could be sold over the counter: this was not a meeting fixed to discuss any side-effects of terfenadine. In October, Merrell Dow requested a postponement of that meeting. On 7 December 1989, Merrell Dow submitted information on cardiac events which the FDA had not previously requested; the next day, the FDA called the meeting which took place in June 1990 and requested a unified submission of information on cardiac effects and interactions. Dr Botstein stated at the meeting that the matter brought to the committee was terfenadine’s ability to cause prolongation of the QT interval and TdP.

Next, reference should be made to Dr McNutt’s account as recorded in the Special Drug Communication Teleconference on 7 August 1992, set out inENT Journal September 1992. He, it will be remembered, worked within Merrell Dow, and says that it was Monahan which had caused Merrell Dow to review their data of concomitant ketoconazole/terfenadine administrations. To the same effect is Woosley II: see paragraph 126 below. As already mentioned, the paper refers to the case reported in Monahan stating that it was this case which prompted the FDA “to review its Spontaneous Reporting System for additional reports of arrhythmias associated with terfenadine (and astemizole)”. This was clearly not a review of the system as a system (ie to assess its effectiveness) but to review the data base for additional reports.

By the time of the FDA meeting, the FDA review of its data-base had shown up a number of cases which were considered at the meeting. It does not seem, however, that it was those cases which were the cause of calling the meeting. I am afraid the details of what actually happened are not before me. But what does seem most likely from what I have seen is that Monahan was the catalyst that caused further searches to be carried out both by Merrell Dow and by the FDA so that in that sense Mr Watson is, I think, correct when he says that it was the Monahan case that led to the FDA meeting to address the cardiac side-effect problem.

Dr Morley relied on a number of papers to cast doubt on the hypothesis that it was unmetabolised terfenadine that was responsible for the QT prolongations which led to isolated cases of TdP. Mr Watson suggested that Dr Morley had really taken a few sentences out of each paper but that they did not really cast any doubt on the hypothesis. I mention most of them, albeit briefly.

Hoppu et al: Accidental astemizole overdose in young children, Lancet 1991; 338: 538-540. This reported six cases of accidental astemizole poisoning in children. They refer to some safety precautions to reduce the risk of accidental ingestion. It is relevant (in the context of class-effect) because of a single sentence:

“Inview of the analogous cardiac effects in terfenadine overdose, similar precautions may also be justified in cases of accidental ingestion of other selective H_i-histamine receptor antagonists.”

Simons: Pharmacokinetic Optimisation of Histamine H₁-Receptor Antagonist Therapy, Clin. Pharmacokinetics 1991; 21: 372-393. This contains an interesting review of the literature to date (including reference to Monahan). It states that patients receiving terfenadine who concurrently ingest ketoconazole may develop elevated serum terfenadine concentrations and cardiotoxicity: that is referenced to a 1991 paper (which I have not seen) by Eller & Okerholm. The wording here might suggest a perceived causality between elevated serum levels of terfenadine and cardiotoxicity. But whether that is a correct inference or not, it is clear that the paper says nothing to suggest that the acid metabolite was responsible to any extent for the adverse cardiac side-effects of terfenadine treatment.

Honig: Changes in the pharmacokinetics and electrocardiographic pharmacodynamics of terfenadine with concomitant administration of erythromycin, Clinical Pharmacology and Therapeutics (1992) 52: 231-238. This paper was submitted for publication in March 1992 and accepted on 2 June. Referring to terfenadine, it records that enthusiasm for this drug has been tempered by a small but increasing number of reports involving a possible association with TdP. It states that “Terfenadine is often thought of as a pro-drug, with the acid metabolite as the active agent…”. The conclusion of the study reported is that it “may support the hypothesis that the accumulation of terfenadine is responsible for the alteration in the cardiac repolarization seen during terfenadine therapy or overdosage. Erythromycin significantly alters terfenadine pharmacokinetics and may cause accumulation of terfenadine to detectable levels in certain patients. It is still unknown what facts are responsible for the individual susceptibility to increased terfenadine concentrations in patients concomitantly receiving erythromycin or ketoconazole. Furthermore, if our hypothesis is correct, the factors that enable high concentrations of terfenadine to lead to torsades de pointes in some individuals but not others are unknown”.

Knowles:Astemizole and terfenadine-induced cardiovascular events, The Canadian Journal of Hospital Pharmacy Vo. 35 No 1 February 1992. This is a drug information note and not what one might call a learned scientific paper. The summary of the report is that astemizole and terfenadine have caused TdP in susceptible patients and warns pharmacists and physicians to be aware that drugs which inhibit the cytochrome P-450 pathways (eg ketoconazole) may inhibit terfenadine clearance and lead to toxicity. Here the focus is on the parent with no mention of any metabolite. Referring to Monahan, the authors remark that terfenadine and metabolite concentrations were remarkably elevated. Dr Morley accepted that the report is neutral on the question. It has to be accepted, I think that this paper contains nothing to give the skilled man the impression that there are serious doubts as to any hypothesis that terfenadine is the cause of TdP.

Davies: Cardiotoxic effect with convulsions in terfenadine overdose. British Medical Journal 1989: 325. See further at paragraph 80 above. This paper pre-dates Monahan and Mathews. I rather agree with Dr Morley here that the note does not discriminate between parent drug and the metabolite since it notes raised concentrations of both terfenadine and the acid metabolite; in that sense it is neutral; it does not support the hypothesis later put forward by Monahan.

Du Buske: Clinical comparison of histamine H₁-receptor antagonist drugs December 1996 J Allergy Clin Immunol 1996; 98:5307-5318. In this article, there is a section dealing with cardiovascular side-effects of H₁-receptor antagonists. It is recorded that among the second-generation antihistamines, only terfenadine and astemizole have been reported to induce QT prolongation. Because there is a link between terfenadine concentrations and the observed increase of QT_c interval, cardiotoxic concentrations of terfenadine can be seen with concomitant administration of ketoconazole, erthromycin and clarithromycin. In both cases, a significant increase in the mean area under the curve of terfenadine carboxylate occurred. It notes that there is some suggestion and that TdP can occur also with large doses of some first-generation antihistamines.

Dr Morley’s observation on this was as follows: “…what this does is it provides evidence that would be considered consistent with the hypothesis that is under discussion but it does not apportion – does not exclude the possibility that metabolites may contribute to this effect”. This paper cites an earlier Honig paper at JAMA 1993; 12:1513-18 which Dr Morley accepted casts no doubt on the perception in May 1992 that terfenadine was the cause of TdP; but notwithstanding that, he was unwilling to accept this in relation to Du Buske. I very much gain the impression that his point was that there was no certainty (something which it is easy to accept) rather than that there was a significant doubt about the hypothesis.

MacConnell & Stanners: Torsades de pointes complicating treatment with terfenadine use, British Medical Journal 302: 1469 (15 Jun 1991). This note reports a case of a woman diagnosed some time after admission to hospital with TdP. For 6 weeks prior to admission, her only drug had been terfenadine. Dr Morley was not really able to say that this cast any serious doubt on the hypothesis that it was terfenadine which was causing the QT prolongation. Professor Frew saw this as an overdose case since the dose before admission was 360 mg daily (significantly above the recommended dose) and saw the paper as a warning not to use terfenadine at that dose.

The note records this:

“TERFENADINE is a specific histamine (H₁) receptor antagonist. It is used by dermatologists and general practitioners at doses up to 360 mg a day….Prolongation of the QT interval has previously been observed in adults taking 120-240 mg of terfenadine a day. Ventricular arrhythmias have also been noted but only in terfenadine overdose or when combined with hepatic enzyme inhibitors such as ketoconazole….

Astemizole, which is also a long acting H₁ antihistamine, has also been reported as causing torsades de pointes when taken in overdose. As there are histamine receptors in the heart, H₁ antihistamine agents may have a direct cardiotoxic effect at higher doses.

Analysis of terfenadine concentrations in our patient showed higher concentrations of metabolites than might be expected in a dose of 360 mg daily. The absence of unmetabolised terfenadine, concomitant drug treatment, and normal results of a liver function test would suggest, however, that her excretion of terfenadine was within the normal range.”

In response to Mr Watson’s question what it was that would indicate to the skilled man that there was serious doubt about the hypothesis that it was terfenadine which was causing the QT prolongation, Dr Morley referred to that last paragraph. He commented that “the suggestion is that metabolites are present there” but did not go further to say that the suggestion was that it was the metabolites which caused the QT prolongation. In any case, this paper would be a report of only one of many reported cases albeit that it would need to be taken into account. As Mr Watson pointed out, the Merrell Dow team must have had this paper when reaching their conclusions.

Zamani: Enantiometric analysis of terfenadine in rat plasma by HPLC, Chirality 1991; 3:467-470. This paper came from the group at the Bethesda Hospital from whence Monahan came. Monahan is cited without demur. It is not possible to read this paper as departing in any way from what Monahan says. This is particularly clear given that two of the authors, Zamani and Cantilena, were co-authors of Honig and that Cantilena was a co-signatory of the response to the Mathews letter (see paragraph 128 below). Referring to terfenadine, the paper states:

“After oral administration, it is rapidly metabolised, presumably by cytochrome P-450 oxidative systems in the liver, to a pharmacologically active carboxylic acid metabolite. Inhibition of terfenadine metabolism by drugs such as ketoconazole, resulting in increased plasma concentrations of parent drug and the carboxylic metabolite, has been associated with significant toxicity [citing Monahan]”

Dr Morley was pushed by Mr Watson to accept that on the basis of the information in the article, it is a reasonable inference that it is the terfenadine which is responsible; it is an inference which a reasonable team would have arrived at in May 1992. He declined to do so. But he did have to accept that nobody by May 1992 was suggesting that the metabolite had any effect which was causing the QT prolongation even though he said it was a matter of opinion whether you attribute the problem to greater concentrations of the parent or the metabolite.

The paper also has a section on the I_k channel, possibly the first report other than Chen of such investigations. Mr Watson put to Dr Morley that, if there is any doubt about the respective roles of terfenadine as against the metabolite, the next stage would be to carry out an I_k channel check. Dr Morley replied “Not for me, but you know, because I am an old-fashioned pharmacologist”. His preference was for an isolated perfused heart preparation, a Langendorff type preparation, because he said that that would pick up more effects than a specialised test. He expressed great scepticism about the test proposed by Mr Watson.

It was not entirely easy to keep Dr Morley focused on the issue at this stage of his cross-examination. He seemed to be more concerned with presenting a different approach, namely pursuing a research programme based on developing a new compound from those in which no QT problems existed, rather than carrying out tests to see whether the acid metabolite had any effect on the I_k channel. He did not see the tests as the last piece in a jigsaw, where the existing pieces were (a) the problem of TdP and (b) the answer to the problem, namely, the hypothesis that it was terfenadine that was being created through hepatic impairment. He saw it as part of a research programme and did not regard a result from the I_k channel test as enough to take that programme along; it was too flimsy a basis for expenditure of many millions of dollars.

Nonetheless, he had to accept that it was a logical step from all the learning that existed in May 1992 to do the appropriate channel blocking test of terfenadine and the metabolite. But he would not agree that the result of such tests – whether the I_k test or the Langendorff heart test – would necessarily be that the metabolite would be shown not to be a channel blocker: the result of the test would all depend on the conditions of the test. This was his view even though those who did the tests got that result and even though the Patents tell us that the tests show that the metabolite did not block the channel. But he did accept – he could hardly have done otherwise – that there was no contrary evidence: see paragraph 214 below.

Indeed, after, as Mr Watson would say ducking and diving about alternative commercial routes, Dr Morley accepted that the team leader, being faced with a decision to give a go-ahead financially, would have the following in front of him or her: the brilliant safety record of the metabolite, the brilliant efficacy of the metabolite, the knowledge that it is highly unlikely that in clinical trials general toxicity would be created, highly unlikely that new side-effects will arise, and almost total certainty that you will have efficacy; and he agreed with all of those. And then came these two questions from Mr Watson:

Dr Morley was also asked some questions about an early paper – Sorkin et al: Terfenadine: a review of its pharmacodynamics properties and therapeutic efficacy, Drugs 1985; 34-56. It notes that a “carboxylic acid metabolite [ie the acid metabolite] has been shown to possess antihistaminic activity and may play a role in the activity of terfenadine in vivo”: this is a reference to a paper by Garteiz et al at Merrell Dow [Pharmacokinetics and Biotransformation Studies of Terfenadine in Man, Arzniem. Forsch. Drug Res 1982; 32: 1185-1190]. At greater length, Sorkin states:

“This suggests that a metabolite, or metabolites, may be responsible for a portion of terfenadine activity in vivo. In fact, while metabolite II possesses no antihistaminic activity, metabolite I has been found to have one-third the activity of terfenadine in guinea-pig ileum (unpublished data by Chen, cited Garteiz et al., 1982)…..”

As to Garteiz itself, the conclusion is as follows:

“In conclusion, these data indicate that terfenadine is well absorbed in man and that in therapeutic doses kinetics are linear. It has also been shown that terfenadine undergoes extensive biotransformation to essentially two metabolic products which are rapidly excreted via urine and feces. One of the metabolic products, a carboxylic acid analog of terfenadine [ie the acid metabolite], has been shown to have antihistaminic activity and may play a role in the activity of the parent drug in vivo.”

There is nothing in the conclusion which suggests that the other metabolic product has any part to play in the activity of the parent drug in vivo and nor do I detect anything in the body of the paper which would lead to that conclusion.

Other papers – Kemp and Woosley I and II below – were produced by Mr Watson.

Kemp et al: Antihistamines – is there anything safe to prescribe? Guest editorial, Annals of Allergy Vol 69 October 1992 pp 276-280. This was an article put to Dr Morley in cross-examination.

Addressing the elevation of terfenadine levels in cases of administration concurrently with ketoconazole and also overdose cases and cases of hepatic impairment, it is noted that there is:

“a strong relationship between the observed increase in QTc and the circulating levels of terfenadine and little or no relationship with levels of the acid metabolite. This suggests that a high level of the terfenadine parent compound is responsible for an increase in QTc and, therefore, the possible catalyst to the development of torsades de pointes.”

Whilst Dr Morley had observations to make about what he called “the rather fixed viewpoint” of the authors reference to “the acid metabolite and its active metabolite” as though the former was not an active metabolite, he did accept that the authors considered that there is a strong relationship between the observed increase in QT and levels of terfenadine and little or no relationship with levels of the metabolite. He also accepted that the skilled man would see a strong relationship between QT prolongation and increases parent terfenadine and little or no relationship with levels of the metabolite

Woosley et al: Mechanisms of the cardiotoxic actions of terfenadine JAMA March 24/31, 1993 Vol 269 No 12 pp 1532-1538 (“Woosley I"). In the abstract under Results and Conclusions it is stated, referring to the 25 reported cases of TdP:

“Predisposing factors in these cases indicated that the parent drug, but not its metabolite, may have actions similar to those of quinidine that are responsible for inducing arrhythmia. In vitro studies found that terfenadine is equipotent to quinidine as a blocker of the delayed rectifier potassium current in isolated feline myocytes. The [acid metabolite] did not inhibit this potassium current even at concentrations 30 times higher than the concentration of terfenadine producing a half-maximal effect.

Since blockade of the potassium channel did not occur with the major metabolite of terfenadine, episodes of torsades de pointes are most likely the result of a quinidine-like action of the parent drug and of factors that impair the normally rapid metabolism of terfenadine….”.

On the basis of Sorkin, Woosley I states that the acid metabolite “is considered to be the active antihistamine”. I do not think that Sorkin goes that far; but by 1993 (albeit after the filing of the Patents) this was the perception of at least some of those active in the field.

Woosley I refers to 25 reports of TdP found in the FDA database through to the end of April 1992. It is stated that 20 of the 25 patients concerned had one or more factors that would be expected to cause excessively high concentrations of the parent drug ie overdose or factors that could impair metabolism.

Woosley: Cardiac Actions of Antihistamines, Annu Rev Pharmacol Toxicol, 1996, 36:233-252 (“Woosley II”). In the Introduction, Woosley states that in June 1990, the FDA convened its Pulmonary-Allergy Drug Advisory Committee to review more than 20 cases from the FDA’s Spontaneous Reporting System of syncope, cardiac arrest, ventricular arrhythmia, or TdP associated with use of terfenadine or astemizole. It might be inferred from that that Professor Woosley – who was on the committee - perceived the reason for the meeting being the cases thrown up by the Spontaneous Reporting System. Later, the paper refers to the case reported in Monahan stating that it was this case which prompted the FDA “to review its Spontaneous Reporting System for additional reports of arrhythmias associated with terfenadine (and astemizole)”.

A number of other papers or reports not already mentioned were referred to in cross-examination of the experts which I should mention.

Mathews et al: Torsades de Pointes occurring in association with terfenadine use, JAMA 1991; 266:2375-6 (November 1991). This was a response to Monahan from the Merrell Dow team, commenting on the report and sharing results from subsequent studies at Merrell Dow. One such study

“confirmed that ketoconazole inhibits the metabolism of terfenadine. High plasma levels of unchanged terfenadine and lower-than-expected levels of acid metabolite were noted. When electrocardiograms from patients who were given terfenadine and ketoconazole were compared with electrocardiograms from patients who were given only ketoconazole, a prolongation of 10 to 20 milliseconds in the corrected QT interval (corrected QT is normally 330 to 470 milliseconds) was noted at times of high, unchanged terfenadine levels and lower-than-normal acid metabolite levels. This concurrence suggests an association between increased levels of unchanged terfenadine and prolongation of the corrected QT interval…”

The debate between the Merrell Dow group and the Monahan group was over whether the patient had taken terfenadine at a dose higher than recommended, the former seeking to throw the blame onto that aspect with the latter denying that there had been any overdosage. But there was agreement that the Monahan hypothesis could be right: the concurrence just noted is identified and nothing at all is said to implicate the acid metabolite. Mathews does not, of course, contain data which confirms the hypothesis.

Eller et al: Pharmacokinetics of Terfenadine in Healthy Elderly Subjects, J Clin Pharmacol. 1992; 332:267-271. This is another paper from Merrell Dow. It repeats the now-familiar fact the terfenadine undergoes extensive first-pass metabolism with more than 99% of the absorbed dose eventually metabolised (referenced to Garteiz). It states that “the major metabolite [ie the acid metabolite] appears to exert most, if not all, of the pharmacologic actions that are associated with administration of the parent drug” (referenced to unpublished data on file, Merrell Dow Research Institute, 1980).

Barbey et al: Cardiovascular Safety of Second-Generation Antihistamines, American Journal of Rhinology 13,235-243, May-June 1999. Professor Frew was a co-author of this paper. Early in the paper appears the following:

“The cardiovascular safety of second generation antihistamines was brought sharply into focus in the late 1980s and early 1990s, when various reports of morbidity and mortality were associated with the use of the nonsedating second-generation antihistamines terfenadine and astemizole. Such adverse events were largely attributable to a rare form of ventricular arrhythmia termed torsades de pointes. The association of such a serious side-effect with the treatment of non-life-threatening illnesses for which these drugs are prescribed raised substantial concerns not only about the specific agents associated with these reports, but also about the entire class of second-generation antihistamines.”

Mr Waugh is keen to establish a class effect because it assists him in his case on obviousness. He of course latched onto the statement in the last sentence not just because of what is said but also because it is said in a paper co-authored by Professor Frew. There was lengthy cross-examination of Professor Frew on this on this.

By 1999, Barbey was able to write that the propensity to block potassium channels and prolong of the QT interval appears not to be a class effect of the nonsedating antihistamines. Rather it is confined to some members of the class, such as terfenadine and astemizole. Under “Patients at Risk”, Barbey includes this:

“Thus, from a clinical standpoint, a decision must be made, first, to prescribe antihistamines and, second, to find the drug with the least propensity for adverse effects. Overall, the second-generation antihistamines are preferred to the first-generation agents because they lack sedative effects. Among the second-generation agents, it is clear that not all drugs are equal. The greatest concern for safety has been caused by reports of serious cardiac arrhythmias like torsades de pointes with the overdosage or drug interaction characteristic of some drugs in this class. Because the blockade of the I_k channel has been identified as the single most important factor responsible for the development of torsades de pointes, it has been said that, in the case of antihistamines, the safest drug is that with the largest gap between the therapeutic dosage and the dosage at which 50% blockage of I_k occurs. Of the second-generation antihistamines, terfenadine and astemizole have been clinically proven to have the greatest potential for cardiac dysrhythmias…”

And the conclusion is that “Overall, the adverse cardiovascular effects of antihistamines appear not to be a hallmark of all second-generation agents. Rather, the cardiotoxic effects of some of the antihistamines appear to be related to their potassium channel blocking affinity….”.

The parties called one expert witness each: Professor Frew for Teva and Dr Morley for the Defendants.

Professor Frew is Professor of Allergy & Respiratory Medicine, Consultant Physician in Allergy, Respiratory Medicine and General (Internal) Medicine, at Brighton & Sussex University Hospitals NHS Trust. He has an impressive cv in hospital and academic posts with a string of contributions to research papers in the relevant field. He has extensive experience as a clinician in the relevant fields.

Mr Watson describes him as ideally placed to educate the Court as to the state of the art at the time and to assist the Court to “don the mantle” of the skilled person. I agree that he had, as Mr Watson says, first hand experience and knowledge of the anti-histamine field at the relevant dates and was familiar with the evolving story of terfenadine and its cardiac side effects.

Unlike Dr Morley, Professor Frew does not have experience working within a large pharmaceutical company selecting drugs for development although he did have experience advising such companies. However, there is no reason to think that his approach as a clinician is any different from that of a clinician called upon to advise at the relevant time. And one might think that the input of the clinician on the skilled team would be likely to be at least as important as that of the pharmacologist.

Mr Waugh was highly critical of Professor Frew’s evidence in a number of respects saying, in summary, that his approach was not that to be expected of an expert doing his best to assist the court. Professor Frew is criticised for showing a marked reluctance to answer questions and instead was prone to give lectures. In relation to that last criticism, it is fair to remark that some of the questions were very long and invited lengthy replies; and it is a criticism which can equally be directed at Dr Morley. In his closing submissions, Mr Waugh gives some examples of the allegedly unsatisfactory evidence which I deal with.

First, Professor Frew did not qualify paragraph 71 of his first report and paragraph 4 of his second report where he said that it was only when high concentrations of terfenadine were present (due to overdose of or drug interactions) that TdP was observed. He should have done so in the light of the 1990 FDA materials (as to the FDA hearing, see at paragraphs 78ff above).

Although it is true that Professor Frew did go off at a tangent when the relevant question was first asked, he did answer it clearly after an intervention from me. He did, it is true, go on to enlarge on the clear short answer in a way which, for my part, I found helpful in elucidating why what he has said in his reports was not precisely correct. The criticism is also a harsh one. Let me set out Mr Watson’s response to it:

“Prof. Frew, quite properly, had written his reports on the basis of his primary knowledge and reading of the literature. The 1990 FDA materials had not been referred to by Dr Morley in any of his three reports and Professor Frew had not been shown them. Professor Frew was then provided with a very large bundle of papers on Friday 20^th while he was seeing patients. He read them over the weekend while travelling to and at his conference in Rome. At no point was he told which passages to focus on or why he was being asked to look at the documents (despite requests made by Teva’s solicitors to the Defendants’ solicitors for clarification). He had not been asked to go back to his reports to check them line by line having read the documents and had not done so.”

I agree with Mr Watson. It may be that Professor Frew was guilty of a departure from perfection, but I do not find in this lapse (whether by itself or taken with his evidence as a whole) behaviour warranting Mr Waugh’s admonitions let alone enough to lead me to reject his evidence.

Mr Waugh refers me to other passages where he says that Professor Frew exhibited his propensity to give speeches and to fail to answer simple questions. With two exceptions, I do not think that these criticisms are fair. Being a highly intelligent and articulate man, Professor Frew often wanted to illustrate that the simple answer to the simple question provided only half the picture. I did not find his answers unhelpful on the whole. The two exception are these. The first is his answers in relation to one of the papers (MacConnell & Stanners – see paragraph 108above) where he seemed unwilling to accept that the paper did not say, impliedly, what he said it did. This concerned histamine receptors in the heart, an aspect I deal with separately at paragraph 213 below. The second is his answers to questions about Sorkin and his failure to address fairly and squarely why there was a reference to a portion in the suggestion that “a metabolite, or metabolites, may be responsible for a portion of terfenadine activity in vivo”.

It is then said that Professor Frew’s approach to the case was unsatisfactory in that

As to paragraph a. (hindsight) it is of course always easy to be wise after the event. However, I have re-read both of Professor Frew’s report and read the entirety of the transcript of his cross-examination. I do not think that, taken overall, his evidence, or more importantly his opinions, are based on hindsight. He has done his best to explain what his view of the state of the art and common general knowledge at the relevant times actually was.

Mr Waugh says that hindsight was introduced into Professor Frew’s evidence from the outset by the way he was instructed. His instructions were to give his views on the assumption that the skilled person would do something. This is particularly important, he says, having regard to the lack of motivation to do anything in the light of the accepted safety of terfenadine and the fact that the TdP was addressed with adequate warnings. I confess to not understanding this point: by the priority date, the cardiac side-effects of terfenadine were cgk and, although labelling could, and did, address the problem, it does not take any imagination to think that a drug as effective as terfenadine but without those side-effects would have an edge over terfenadine. Indeed, once the acid metabolite had obtained its licence and been marketed, terfenadine ceased to be available for treatment of allergic rhinitis and urticaria.

As to paragraph b.i. (cardiac events only seen with excess terfenadine) this is the issue which I have already dealt with in relation to paragraph 71 of his first report and paragraph 4 of his second report.

As to paragraphs b.ii. and iii., these are dealt with respectively at paragraphs 131ff and 95ff above. My conclusions are that there was no perception of class effect and that the Monahan case (rather than the paper) was the trigger for the FDA meeting; at least, it accounted for the agenda for and focus of that meeting.

Accordingly, I do not think that the position taken by Professor Frew on these aspects undermines in any way the evidence which he gave.

My own assessment is that Professor Frew was an impressive witness who gave his evidence fairly, but who forcefully defended his views against Mr Waugh’s sustained questioning.

Dr Morley also has an impressive cv and has great experience as a pharmacologist. He has extensive experience of histamine and allergic reactions. He has also worked in the pharmaceutical industry; he was Head of Asthma Biology and Preclinical research and Development at Sandoz Pharma from 1984 to 1992.

However, he had no interest at the time, up to the priority date, about the emerging events relating to the QT problem (the cardiac side-effects) of terfenadine. At the relevant time, he worked for Sandoz and, as Dr Morley said, there was no particular interest in Sandoz in the mid to late-1980’s with cardiac events in this field. He had read, at best, very few of the key documents when they were published. As he accepted, he had constructed his view of the state of the art about cardiac issues from reading the documents presented for the purposes of the present case. I must, and do, of course take that absence of direct knowledge into account in assessing his evidence about the state of the cgk and people’s thinking and attitudes, accepting, as I do, that he is not disqualified from expressing expert views: an expert is allowed to read his way into a case as Pumfrey J explained in Research in Motion UK Ltd v Inpro Licensing Sarl [2006] RPC 20 at paragraph 6. Moreover, the skilled team in the present case is a team within a pharmaceutical company; Dr Morley has the advantage of experience of working in such an environment which Professor Frew lacks (other than as an external consultant).

I mention here a number of criticisms which Mr Watson makes.

The matters referred to in b, c and d are, I think, factually correct. Whether they are properly to be regarded as criticisms or not, I rather doubt. I take account of the points. The first criticism will be covered in later parts of this judgment.

My own view is that Dr Morley, like Professor Frew, was an impressive witness doing his best to help the court. That he and Professor Frew did not agree on everything reflects, in large part, their different perspectives and experience. I found them both helpful. Their function, of course, is to educate the court; and although they are entitled to express their own opinion on the ultimate question of obviousness, my function is not to choose between the experts on the correct answer. The question is one for me to answer in the light of the totality of the evidence.

Ignoring Claim 12 of 443, the alleged invention in the present case is using the acid metabolite to make a medicament for use as an antihistamine and to treat allergic rhinitis and urticaria. Mr Watson says that this is disclosed in Carr which is relied on as prior art. This is a clear case, he says, of anticipation over Carr and a blatant attempt to re-patent the prior art.

Mr Watson no longer pursues the case of lack of novelty of the Sepracor patents based on the 976 application. Nor, as I understand it, does he rely on the 1992 PDR for this purpose. I confess that, in reviewing all the material in preparing this judgment, I have been unable to find anything in his written or oral submissions where he actually says as much, but Mr Waugh presented his case on that basis, expressly deleting the parts of his closing submission relating to the point, Mr Watson making no comment to suggest that this was wrong. I therefore deal only with Carr.

The relevant disclosures in Carr are set out at paragraph 68above. To recap, Carr discloses:

The issue of novelty needs to be addressed in the context of conventional novelty and in relation to Bristol-Myers. Mr Waugh makes two points in relation to conventional novelty.

The first is that there is not the required clear and unambiguous disclosure in Carr that the acid metabolite is to be used as a pharmaceutical composition to provide an antihistaminic effect. This applies generally to claims 5, 9 and 12 of 976, claims 1 and 12 of 443 and claims 1 and 2 of 860.

The second point is that, in relation to claim 12 of 976, claim 1 of 443 and claim 3 of 860, novelty is conferred also by virtue of the manufacture for use to treat allergic rhinitis and urticaria. Mr Waugh says that there is not the required clear and unambiguous disclosure in Carr of the use of the compounds (in particular of the acid metabolite) for the treatment of allergic rhinitis and urticaria. He says, contrary to Mr Watson, that “allergic rhinitis is a condition which is distinct from merely an antihistamine”. Or, to put it another way, the discovery that it is useful to treat these conditions meets the criteria for novelty required by Bristol-Myers. He submits that a mere disclosure that something is or might be useful as an antihistamine is not a disclosure of the use of the compound to treat a specific indication such as allergic rhinitis. In practical and regulatory terms, such indications are treated distinctly.

As to the first point (no clear and unambiguous disclosure) in relation to 976, it is said by Mr Waugh, so far as concerns claims 5 and 9, that Carr does not make a clear and unambiguous disclosure that the acid metabolite is to be used in a pharmaceutical composition to provide an antihistaminic effect. He relies on certain passages from Dr Morley’s first report as well as on answers from Professor Frew in cross-examination. The material from Dr Morley’s report is found at paragraph 49 (which I set out below in relation to Mr Waugh’s second point) and paragraphs 100-107. I would also add paragraphs 57 to 60 as to which also see below.

In those passages, Dr Morley identifies the three uses of the compounds disclosed in Carr namely anti-histamine, anti-allergy and bronchodilators. He explains that the skilled addressee would not have thought that any one of the compounds disclosed by Carr would be likely to have all three of the properties referred to, something which Mr Waugh says was not challenged. Although it may not have been directly challenged, Professor Frew was asked about it and did not accept the proposition. Dr Morley also considered that the guinea pig assay is not specific for histamine H₁ antagonists, for instance the bronchodilator Salbutamol can produce a positive result in the ileum test but has no histamine H₁ antagonist activity; it is not an anti-histamine.

Dr Morley said that drugs used to treat allergic rhinitis at that time, 1981, were administered by tablet or oral suspensions and it was well known (excepting one specified condition) that aerosols were ineffective in this application. Accordingly, Mr Waugh submits that Carr really suggests the use of the acid metabolite as a bronchodilator for the treatment of asthma: this is on the basis of Example 12 which shows the acid metabolite as an illustrative composition for an aerosol suspension. It is certainly the case that no example of an antihistamine spray was identified; accordingly, according to Mr Waugh, the skilled reader of Carr would take the example of the proposal to use the acid metabolite as a of bronchodilator and not as an antihistamine.

There is an important passage at the end of Dr Morley’s cross examination which I should set out. To put this in context, it is to be noted that in his first report Dr Morley sets out Carr as suggesting that the compounds are useful as antihistamines, anti-allergy agents and/or bronchodilators, whereas it actually said “and” rather than “and/or”. This is not as trivial as it might seem because of Mr Waugh’s argument that the use of the acid metabolite as an antihistamine is simply not disclosed. The interchange between Mr Watson and Dr Morley went as follows:

“A. Yes, the only reason I questioned it was I know of no compound in this category, whether antihistamine -- compounds that are antihistamine compounds that also are significant anti-allergy agents and significant bronchodilators. So you have a classification which is academically interesting, but meaningless.

Q. Right. Carr does not say significant, does he?

A. No.

Q. You would expect, based on the fact that it is closely related to terfenadine, that the novel compounds would at least have antihistamine activity, would you not?

A. Yes.

Q. The guinea pig ileum test I think we arrived at probably where you would agree. It is the standard test to indicate whether a compound has antihistamine activity, but it does not of itself prove that it is an antihistamine. Would that be fair?

A. Yes.

Q. Now, if you are told something is useful as an antihistamine, that would immediately suggest allergic rhinitis, would it not? Can you think of anything more mainstream?

A. If it was sufficiently potent, definitely.

Q. ….. When you said sufficiently potent, does allergic rhinitis require greater potency than any other antihistamine activity?

A. No; I meant sufficiently potent to make it worthwhile to give to people. If it is not very potent, you have to give very large amounts of it, and then its prospects of commercial success are very slight.

Q. That is fair enough…... If it is in truth an antihistamine as Carr tells you, it will give some relief to allergic rhinitis. How good you would have to test. Is that fair?

A. Yes. If it is antihistamine, if it has these other properties, you have to then decide do they complicate it, do they preclude its use. If it is a simple antihistamine, it is just I think really a matter of is it sufficiently potent and selective to use it.

Q. Yes. It may not be a drug which you are going to bet the company's finances on, but if Carr is correct and these compounds, including the metabolite, are antihistamines, you would anticipate and expect that it would have therapy against allergic rhinitis?

A. It could have, if you developed it, yes.

Q. I mean, if it is an antihistamine, it is highly unlikely to have no effect on allergic rhinitis, would you agree?

A. Yes.”

As to the question about significance asked by Mr Watson in that exchange, Mr Waugh observes that, as a practical document, the skilled reader is not concerned with insignificant effects. Unless the activity is significant the compound will not be useful. Dr Morley accepted that if told that the compound was a potent antihistamine then it would suggest use for allergic rhinitis. However neither potency nor use for allergic rhinitis is actually disclosed by Carr. Mr Waugh says that Carr in fact provides very little information as to the properties of the acid metabolite and adopted Dr Morley’s observations about the guinea pig assay (ie that the test is not conclusive that the acid metabolite has anti-histaminic properties). He reiterated that the aerosol formulation in Example 12 suggests use as a bronchodilator.

As to the second point (novelty conferred by manufacture for use to treat allergic rhinitis) Mr Waugh submits that there is no unambiguous disclosure of the use of the compounds (or the acid metabolite) for the treatment of allergic rhinitis. He relies on Dr Morley. It is important to see what Dr Morley actually says in paragraph 49 of his first report which is this:

“49.

Early histamine (H₁) receptor antagonists or anti-histamines, i.e. those which were commonly available by the mid 1970’s, included the ethanolamines, alkylamines, ethylenediamines, piperazines and phenothiazines. Clinical comparison of these various compounds revealed differences that could not be predicted from their potencies as histamine (H₁) antagonists. These differences reflected other pharmacological properties such as antagonism of muscarinic receptors, serotonin receptors, a₁-adrenoceptors, as well as local anaesthetic action and suppression of the release and formation of biologically active materials during allergic reactions. These early histamine (H₁) antagonists became established during the 1960’s and 1970’s as therapies for hay fever (e.g. chlorpheniramine, clemastine) for motion sickness (e.g. promethazine, diphenhydramine, cyclizine and cinnarizine), as sedatives (e.g. promethazine) or for certain forms of dizziness (e.g. dimenhydrinate, cyclizine). However, compounds which were useful in allergic rhinitis were not necessarily useful in treating other allergic conditions, and vice versa. By 1981 all of this would have been commonly known to the skilled team.”

Mr Waugh relied on what Dr Morley says in paragraph 57 of his first report; all of paragraphs 57 and 60 are of help and I think it would be helpful to set out paragraphs 57 and 60:

“57.

By 3 March 1981, a research team skilled in the art would have been aware that compounds selected as histamine antagonists might be effective in suppressing some of the symptoms of allergic disease, especially allergic rhinitis (hay fever), and might or might not also be effective as anti-allergy compounds and/or bronchodilators.

……..

60.

In summary, by 3 March 1981, a number of selective histamine (H₁) receptor antagonists, including terfenadine, were known as anti-histamines. They were not known to have significant effects as anti-allergy compounds generally or as bronchodilators, while known bronchodilators such as salbutamol were known not to be effective against allergic rhinitis (hay fever). Thus in 1981 it was known that while anti-histamine, anti-allergy agents and bronchodilators were related, there was no certainty or pattern as to the effectiveness of any particular compound; i.e. whether a compound which was useful for one of these uses would be useful for any of the other uses.”

These aspects make it clear, he submits, that Carr is not a novelty destroying reference.

Professor Frew did not agree with the proposition that the skilled addressee would not have thought that any one of the compounds disclosed by Carr would be likely to have all three properties (antihistamine, anti-allergy, bronchodilator). Thus, after a long question and a long answer, there occurred this interchange:

Professor Frew did accept, however, that antihistamines in general are not good bronchodilators and do not work in clinical asthma. He also accepted certain limitations about Carr:

“Q: You cannot predict on the basis of this document which of these particular compounds, if you took the list on column 4, for example, which would have which property, which would be antihistamine which would have anti-allergic properties and which would be effective, if at all, as a bronchodilator.

A: I certainly could not, on the basis of the structures, predict that, no. What it could test though is knowing that a similar structure, the one that is excluded from the abstract, is an effective antihistamine, ie terfenadine, one could predict that some of these structures would be effective as antihistamines. It may not be all, but it would seem quite likely that some of these would be effective antihistamines. No doubt that is the reason why people pursued this and then present evidence further on in the patent which suggests that they are.”

But that answer has to be taken in the context of other answers which he gave, particularly in relation to his view that the guinea pig ileum test, whilst not confirming in absolute terms antihistamine activity (since it is a test which some non-antihistamines would pass) strongly suggests antihistamine activity. So at least there is a reason to think that the acid metabolite was one of the compounds which did have anti-histaminic activity.

Mr Watson submits that Carr quite clearly does disclose the use of the acid metabolite as an antihistamine and does so with the necessary clarity and lack of ambiguity. He makes the forensic point that disclosure is acknowledged in 976 itself. I remind myself that Carr must be construed as of 1981: whilst it was generally accepted in 1992 that the acid metabolite was produced on administration of terfenadine and was, at the least, believed to make a significant contribution to its efficacy as an anti-histamine that was not cgk in 1981. But that does not alter the fact that 976 says that it was Carr which disclosed the use of the acid metabolite as an antihistamine rather than that the compound has since been discovered to be an antihistamine.

In my judgment, Mr Watson is correct. There are two places in Carr where the use of the compounds is referred to. Under Field of Invention it is stated that the invention relates to compounds which are useful as antihistamines, antiallergy agents and bronchodilators; and in the Detailed Description it refers to compounds of the invention being useful for those same purpose. In neither case is any distinction drawn between any of the compounds, relating them to one (or more) of the particular purposes. Further, in the second place, it is stated that the compounds “can be in solid or liquid form such as, tablets, capsules, powders, solutions, suspension or emulsions”. Again, there is no distinction between one compound and another. On a literal reading, therefore, each compound is, potentially, capable of performing each purpose in each method of delivery.

I do not think that the skilled addressee would have read Carr in any different way as restricting any particular compound to any particular purpose. In particular, I do not consider that example 12 entails that the compound concerned (in fact the acid metabolite) was restricted to one to be delivered by that method or that its primary purpose must be taken to be that which it was cgk would be delivered by that method ie a bronchodilator rather than an antihistamine or anti-allergy agent. In my judgment, the teaching of Carr includes that the acid metabolite is a useful antihistamine which can be administered orally. There may be much other teaching in Carr but that does not detract from that teaching.

Now, it may be that the teaching is not enough to satisfy the skilled reader that any one of the compounds is effective for any particular purpose. In particular, he may not be sure that the acid metabolite would in fact prove to be a useful antihistamine in clinical practice. I do not see why that is relevant. After all, the acid metabolite is, as Mr Watson says, a prominent member of the class of compounds and the only one to have been tested in any way. If the teaching of Carr is implemented, it will, not simply might, result in something which would infringe each of the relevant claims in each of the Patents other than, perhaps, claims specifically related to seasonal rhinitis and urticaria, and also claim 12 of 443, to which I will come in a moment.

Professor Frew accepted that it would be necessary to show adequate potency and selectivity as a histamine receptor antagonist in order to use the acid metabolite for a specific purpose, for it to be therapeutically useful or to propose its use therapeutically. But that is not, I think, a precondition to patentability and, indeed, the Patents in Suit are no further advanced than Carr in that respect.

I do not consider that the conclusion which I have reached about the teaching of Carr is affected by the need to show adequate potency and selectivity as a histamine receptor antagonist to establish a clinical use. At least, no convincing reason has been given to me why it should. In any case, none of the Patents show the potency and selectivity and it is common ground how little characterised the pharmacology of the acid metabolite was. That does not answer the question which I have to answer, but it does point up some difficulties for the Defendants if Mr Waugh is right to rely on this point.

Nor am I persuaded by Mr Waugh’s argument that it is not enough to defeat the novelty of any of the claims to show that on the basis of the disclosure in Carr the skilled person would expect (whether on the basis of similarity with terfenadine or otherwise) that the novel compounds would at least have antihistamine activity; such matter must be explicitly disclosed which it was not. It seems to me that that is too narrow a view of the teaching of Carr. My view is that the teaching of Carr as appreciated by the skilled addressee is that the acid metabolite does have antihistamine activity not simply that the skilled person would expect it to have such activity.

In any case, I accept Professor Frew’s view that the guinea pig ileum test, whilst not confirming in absolute terms antihistamine activity strongly suggests antihistamine activity. So, although one would not know that the acid metabolite was in fact effective as an antihistamine, one would have a good reason for thinking that it was and be able to carry out more focused experiments to confirm or refute that hypothesis. Accordingly, the case is not one of the skilled addressee having an expectation of a result which is not disclosed; rather it is a case of an actual disclosure of the result supported by material which suggests that that result is indeed the case.

Carr clearly discloses, I consider, the method of making the acid metabolite: see example 3. Further, I consider that it clearly discloses the use of the compounds generally to make medicaments, as shown by examples 10 to 13 (which relate to the making of tablets, aerosol solution, aerosol suspensions and injectable suspension respectively). These are reflected in claims 10 and 11.

What Carr does not expressly and separately disclose is the use of the compounds to manufacture a compound to treat allergic rhinitis or urticaria. Accordingly, claim 12 of 976, claim 1 of 443 and claim 3 of 860 stand on a different footing from claims 5 and 9 of 976 and claim 1 of 860.

It is established, I think, that an antihistamine will be useful in the treatment of allergic rhinitis and urticaria. That was cgk in 1981 even if, as Dr Morley contends, not all antihistamines were useful for all allergic conditions. 976 itself again acknowledges this (at least as of 1992) where it says in paragraph [0010] that certain compounds (which include the acid metabolite) “are known H₁-receptor antagonists and as such provide relief of symptoms associated with histamine-mediated diseases and conditions such as seasonal allergic rhinitis, urticaria and the like”. Dr Morley acknowledges, in the passage from his cross-examination which I have set out above, that, if the compound was an antihistamine, it would be highly unlikely to have no effect on allergic rhinitis. And Professor Frew, in his first report, identified antihistamines as being useful in both allergic and non-allergic forms of rhinitis and urticaria.

In these circumstances, I do not consider that it is a new form of therapy to use the drug as a treatment for allergic rhinitis or urticaria as compared with its use as an antihistamine in applying the Bristol-Myers approach to second medical use, an approach which is, for reasons already given, binding on me.

Mr Waugh relies on the integers of the claims which relate to the avoidance of the cardiac side effects seen with terfenadine (claims 9 and 12 of 976, claims 1 and 12 of 443 and claims 1 and 3 of 860). He says that the claims are not mere claims to discoveries as such but “relate to new methods of manufacture of a medicament using a compound the acid metabolite for treating the claimed indications whilst avoiding the cardiac side effects seen with terfenadine”. The claims are not, however, directed at new methods of manufacture; the novelty, if any, must lie, in accordance with Bristol-Myers, in a novely therapy. I do not consider that the integers of the claims relating to cardiac side effects produce the necessary novelty if it is otherwise missing. I agree with Mr Watson that the fact that administering the acid metabolite avoids the side effects of terfenadine is a mere discovery about the old use proposed by Carr. The contrary argument was rejected in Bristol-Myers and has also been rejected by the EPO in T1031/00 (Sepracor) at paragraph 2.3.

Accordingly, the following claims are invalid as lacking novelty over Carr: claims 5, 9 and 12 of 976, claim 1 of 443 and claims 1 and 3 of 860.

In contrast, Mr Watson accepts that claim 12 of 443 (which is not in Swiss form) has bare novelty because Carr discloses 50 mg compositions whereas this claim discloses a 60 mg composition. I will return to this in the context of obviousness.

Teva’s position as eventually formulated in Mr Watson’s closing submission is that the case continues (as in opening) to be a straightforward case of obviousness based on cgk alone and/or taken with Chen. There is a slightly different approach based on Carr which I will come to later.

It is to be noted that the technical content of the Patents themselves is in fact quite limited:

On that basis, Mr Watson submits that this is going no further than that which would have been obvious at the priority date; so that the further clinical testing which Dr Morley referred to as being necessary to bring the drug to market is irrelevant to the issues in this case. Mr Waugh’s response to submissions along those lines is that the practical reality is that these matters can only be assessed through extensive clinical trials. Such trials can only be undertaken once a patent application is in place, since it is almost inevitable that clinical trials on such a scale will put the claimed concept into the public domain. For this reason the remedy, should the patentee’s prediction prove unfounded, is an allegation of insufficiency not obviousness. But that, as I see it, is not really to meet Mr Watson’s point: he proceeds on the basis that the Patents are sufficiently described and his simple point is that the description does not go beyond what is obvious from the cgk and Chen.

In addressing the question of obviousness, the problem/solution approach can be of assistance. In the present case, the problem was that patients with hepatic impairment or using certain medications such as ketoconazole could experience QT prolongation and, in extreme cases, TdP even leading to death. The problem had, in practical terms, been addressed in 1990 by the re-labelling and important drug warnings which I have already dealt with. But the underlying problem remained and questions therefore hovered in the air about dealing with it: Was it desirable, or even necessary, to deal with and if so was there an obvious way (or even more than one way) of dealing with it. There can be no doubt that it was a desirable goal to achieve a solution: as Dr Morley agreed, “terfenadine without the torsades risk would be highly desirable”. There was a commercial opportunity. The only real question is whether it was obvious, in the light of the cgk as of the priority dates of the Patents (alone or with Chen) to use the acid metabolite for making a medicament for producing an antihistaminic effect or as a treatment for allergic rhinitis and urticaria.

Whether or not the contents of the papers to which I have referred are included in the cgk, it is now common ground that the skilled team would obtain all of the publicly available information about the cardiac side-effects of terfenadine. Dr Morley did at one time express the view that Monahan would not form part of the literature which the team leader would obtain but that is no longer the submission of Mr Waugh. That, at least, is as I understand the parties’ positions but, if it is not, it is one which I think is clearly correct. What, then, was the state of the art concerning cardiac side-effects – the problem to be addressed – in 1992?

There was in the public arena at least the hypothesis that it was the parent drug which was the cause of TdP. This is clear from Monahan in the passages which I have quoted at paragraph 60 above. In particular:

“The marked increase in the serum concentration of the parent compound may have produced the arrhythmia seen in this case. Identical cardiotoxicity has been reported with acute terfenadine overdosage. The presence of a relatively low serum concentration of metabolite supports this hypothesis.”

The skilled team would read Monahan and appreciate this important aspect of the paper: this hypothesis was not contained as a minor aside in a footnote.

All the same, Mr Waugh is, I consider, clearly correct when he says that Monahan does not provide clear evidence that the incidence of cardiac side effects seen on terfenadine administration were solely caused by terfenadine itself and not by the acid metabolite. However, although the acid metabolite is not expressly exculpated, I think that the only natural reading of what is said, and the reading which the skilled person would take away, is that the reason why there is cardiotoxicity is because there is a high level of parent terfenadine in the serum (and it is there because of an interaction between terfenadine and ketoconazole); the reason is not that there is acid metabolite present.

I do not agree that the skilled person reading Monahan would take the view that the acid metabolite was implicated in the cardiac events seen on the administration of terfenadine. I acknowledge what Dr Morley has to say about this in paragraph 14 of his second report:

“The acid metabolite was a potential candidate for adverse side effects since large quantities of this metabolite are present in plasma for long periods (half-life of 12-14 hr). Protracted exposure implies that the metabolite may be causing cardiac side effects since onset of torsades de pointes was not detected until approximately 10-15 hours after the last administration of terfenadine, when little or no terfenadine is found in the plasma. For example, the patient described by Monahan experienced torsades de pointes 10 hours after all medications had been discontinued following admission to hospital.”

But the acid metabolite (in contrast with parent terfenadine) is present in plasma in ordinary cases where there is no overdose, hepatic impairment or concomitant therapy. What Dr Morley says does not, I think, justify the conclusion that the skilled person would take the view that acid metabolite was in fact implicated. I accept that it was a possibility which could not be dismissed. But the hypothesis that parent terfenadine was the cause was, I consider, clearly stated. Accordingly, Monahan at least provides a pointer in the direction of the parent drug as the culprit and provides no positive indication that the acid metabolite might be implicated.

This is consistent with the approach of the FDA committee. There was a debate where different possibilities were raised, with Dr Lazzara considering that the evidence pointed to the acid metabolite but with Dr Burke immediately raising a question mark over that. Dr Woosley took a different view. He says in relation to his view “That is my guess. That is all we can do right now”. This is an American speaking and when he says “guess” he does not mean that he is sticking his finger in the air (as I rather understand Mr Waugh to say); it is a manner of speaking which conveys the message that this is what Dr Woosley thinks but cannot by any means be certain.

Monahan – the actual paper – came out, of course, some months after the FDA committee meeting in June 1990. Whatever account the committee may have taken of the case as part of its review, the Monahan team was by then willing to speculate quite firmly that parent terfenadine in the serum was the cause of the QT interval prolongation leading to TdP; and it was that paper which passed into the public domain and was influential in the forming of views.

To similar, but stronger, effect as Monahan is Mathews: see the passage which I have quoted at paragraph 111 above. Dr Mathews, as it will be remembered, was a Merrell Dow man.

It was suggested to Professor Frew that the levels of the acid metabolite, as well as of parent terfenadine, found in the Monahan patient were also higher than those recorded in the paper as being expected. But as Mathews shows, the levels of acid metabolite were not unusual: this was a point made by Professor Frew in cross-examination and which I accept, that perceptions about what was a normal level had changed as a result of available clinical data by 1992 so that, by the priority date, the skilled team would not have perceived the Monahan level as out of the norm. Perhaps more significantly for present purposes, the point is that Monahan itself shows the levels of the parent drug to be disproportionately high as compared with the (then perceived) heightened level of the acid metabolite.

There is another issue raised by Dr Morley in relation to Monahan. Monahan, he says, had not addressed the possibility that the metabolite was in some way responsible for the cardiac effects. The study was not he notes of an effect of terfenadine on a patient. Rather, it is an effect of terfenadine in the presence of ketoconazole and the authors simply make an assumption that ketoconazole has no effect on the cardiac action potential. However, as I read Monahan, the authors did consider the metabolite and their view was that it was the parent drug that was responsible. Thus in the commentary (referred to at paragraph 60 above) they say this:

“We believe the serum concentrations of the parent drug and the main metabolite are the result of the inhibition of terfenadine metabolism. The marked increase in the serum concentration of the parent compound may have produced the arrhythmia seen in this case. Identical cardiotoxicity has been reported with acute terfenadine overdosage. The presence of a relatively low serum concentration of metabolite supports this hypothesis.”

As to the suggested assumption being made (and the suggestion that the contrary possibility should have been considered), it is to be noted that Mathews, produced by the men and women from Merrell Dow, does not challenge the assumption. Indeed, the passage from Mathews quoted at paragraph 132 above supports the conclusion that there is an association between increased levels of unchanged terfenadine and prolongation of the corrected QT interval.

In any event, Dr Morley’s evidence was that knowledge that ketoconazole could affect the action potential came later and after the priority date. I was not presented withany material to show that it was known by the priority dates that ketoconazole had any effect on its own on the action potential or the QT interval. It was not a matter which was put to Professor Frew.

Whatever Dr Morley’s final views on this aspect were, Monahan and Mathews established, in my judgment, a simple and clear hypothesis namely that it was the parent drug which, in certain conditions, was the cause of the cardiac effects.

Quite apart from Monahan and Mathews, by the priority date the skilled team would, as is common ground, have had access to and taken account of all the available materials. Thus they would have known about the 25 TdP cases reported in Woosley I at p 1534 being reports of US cases to the FDA to the end of April 1992. As stated there, 20 out of those 25 cases had factors that would be expected to cause excessively high concentrations of the parent drug. Mr Watson suggested to Dr Morley that this would be further confirmation that it was the parent drug which was causing the TdP. Dr Morley disagreed, saying that it was a matter of conjecture that the parent drug was implicated to the exclusion of the metabolite. There was this interchange with Mr Watson:

“Q: ….Where does anybody every say that it could be the metabolite?

A: Well, the metabolites do not go away. When you observe a situation where you have a drug and its metabolite, they are both present.

Q: ……Isn’t the answer, to the best of your recollection, nobody says that it could be the metabolite?

A: It is not something that people have been considering. Some people have not been considering at that time.

Q: No, so nobody raised it.

A: But if you have to read this in a different context in the sense of looking for drug discovery, then you have to change your approach. You have to consider these possibilities. You must not just reject metabolites and say they are mere metabolites, they do not [do] anything; that is not a reasonable approach.”

Then, in answer to a question from me about how strong his doubts might be – the fact that you do not rule something out does not mean that you do not hold a strong view in the same direction – Dr Morley said that in most circumstances he could take a view that would rule out the metabolite. But he considered that the pharmacology available on the acid metabolite was so slight that he could not do that. I can accept that answer as far as it goes. But it does not detract from the fact that the pointers were in that direction or explain to me why the caution which Dr Morley expresses as a pharmacologist should preclude carrying out a test to eliminate the lingering doubt.

To the same effect was this exchange in the course of cross-examination about Kemp:

“Q: That the metabolite had any effect that was creating the QT. Nobody suggested that.

A: They would not know.

Q: No, but nobody suggested it. Everybody suggested it was the terfenadine. What I am saying is that nobody suggested it was the metabolite.

A: Yes.”

It is said by Mr Waugh (and by Dr Morley) that Teva puts far too much emphasis on Monahan and Mathews. He says that Monahan was only one of several reports and was not prominent in the minds of those involved in the area. It is as part of that criticism that Mr Waugh was keen to establish that it was not Monahan which was the trigger for the FDA meeting. Further, he places reliance on the cases referred to in the presentations to that committee to show that the Monahan hypothesis, far from being established, was inconsistent with some of those results. He also relied, to the same effect, on Davies as the first report in the UK of adverse cardiac effect from terfenadine (an overdose case). However, Teva’s case is that by 1992 further data was available and the cgk had moved on (although I note here that this additional material, according to Mr Waugh, did not form part of Teva’s original case or the pleadings, an aspect I will need to deal with).

I have already mentioned the relevant contemporaneous papers and have made various observations in relation to them: Honig, Simons, Woosley I. I agree with Mr Watson when he says that nothing in those papers casts any doubt on the hypothesis which one finds in Monahan and Mathews. It is also important to note in this context what Dr Morley had to say in his cross-examination in relation to Kemp (see the observations at paragraph 125ff above) where he accepted that the authors’ considered that there is a strong relationship between the observed increase in QT and levels of terfenadine and little or no relationship with levels of the metabolite. Although Kemp was not published until October 1992, all the references (which include Monahan, Mathews, Honig, MacConnell & Stanners, Snook and Simons) all pre-date the priority date and these authors, at least, clearly formed the views which they did on such material. It was not suggested by Dr Morley that the cgk had moved on in any relevant way between the priority date and the date of this paper.

Importantly, Dr Morley also accepted that the skilled man would see a strong relationship between QT and increases in parent terfenadine and little or no relationship with levels of the metabolite.

As to Professor Frew, his evidence in his first report was that it was clear by 1992 that the cardiac effects of terfenadine were due to the parent drug and not induced by the acid metabolite. He was cross-examined at length on this conclusion and the material which led to it. But he was not shaken in his view; and I am bound to say that I found the reasons which he gave convincing. [Without setting them out, the relevant passages are set out in Mr Watson’s closing submissions at paragraph 67 and can be found at transcript day 1 p 163 lines 9-32 and day 2 p 254 line 4 to 256 line 7.]

Once it is accepted that the Monahan/Mathews hypothesis was supported by the material available in May 1992 and that there was no suggestion that the acid metabolite played a part in the adverse cardiac side-effects of terfenadine, the evidence also established, in my judgment, that it would have been obvious to carry out further (and well-established) tests. In particular, it would have been obvious to test both the acid metabolite as well as terfenadine itself for their effects on I_k channel blocking and QT intervals. This would be done in order to allay such lingering doubt as there might be over whether the real cause of the problem was the parent drug. It is not only the evidence which I have referred to which points strongly to adopting that course of conduct, but it is also the case that there would be nothing novel or surprising in doing so. The use of a metabolite to avoid the side-effects of a parent drug would not have been novel. As Professor Frew explained, and Dr Morley agreed, this had been done with cetirizine and hydroxyzine. Of course, the fact that use of a metabolite in one case avoids an adverse side-effect of one parent drug does not mean that the use of a metabolite in another case will avoid a different side effect of a different parent drug. But the idea is not new. It is not as though someone could have come along in 1992 and said “I’ve had this brilliant idea of using the metabolite of a drug to avoid a side-effect. Nobody has ever had such an idea before. Let’s do some tests to see if it works”.

I need to consider the suggestion made in MacConnell & Stanners, (see the second paragraph quoted at paragraph 109 above) of a possible direct effect of H₁ antihistamine agents on H₁ receptors in the heart. This was not picked up at any time by anyone else, as Dr Morley accepted. Professor Frew said that he did not find this suggestion persuasive and did not believe that the skilled person would have done so either. Dr Morley, in paragraph 3 of his third report, disagreed with Professor Frew. He relied, in the second half of that paragraph, particularly on a paper by Wolff & Levi (which I do not think I need to attempt to summarise) saying, on the basis of that paper, “Hence it was not unreasonable to consider the possibility that some cases of arrhythmia in humans might stem from interaction of histamine H₁ receptors with histamine or other compounds such as anti-histamines that act at these receptors”. After hearing the cross-examination of Professor Frew and Dr Morley on this aspect, I do not think that Wolff & Levi really support the proposition from Dr Morley just quoted. Nor do I find persuasive the more general statements in the first part of paragraph 3 of the third report that the skilled person would be convinced by the suggestion made in the report that the cardiac effects of terfenadine may be linked to the presence of H₁ receptors in the heart.

I have already described the I_k channel and its blocking and a test which could be used in relation to the channel-blocking effect of a compound such as the acid metabolite. Dr Morley had a preference for a different test – the Langendorff heart test – because he thought that the I_k test could be manipulated by choice of conditions to give the desired effect. But it remains the case that the I_k test was at the time a recognised test and one that would have been known to the skilled team (since the team would include an electrophysiologist). Dr Morley says in his reports that it was not a test which was in any way established for antihistamine compounds; nor was it a test which was validated for pharmacology very much and certainly not in antihistamine pharmacology. However, in parts of his cross-examination which I have already dealt with he accepted that it was a logical step from all the learning that existed in May 1992 to do the appropriate channel blocking test of terfenadine and the metabolite: see the observations onZamani at paragraph 115 above.

As to manipulation of the Chen test, Mr Watson points out, correctly, that the idea that it could occur came from publications after the priority date and could, therefore, have had no impact on the skilled man. Dr Morley’s approach was not put to Professor Frew in cross-examination.

In his second report, Professor Frew says that Chen provides a physiological explanation for the observed link between high concentrations of terfenadine and TdP and provides further confirmation that it was unmetabolised terfenadine that was causing the TdP. Mr Watson says that if one wanted to remove any doubt, it would have been obvious to subject the acid metabolite to the Chen test. This would show that the acid metabolite did not block I_k (and see example 3 of 443). Dr Morley disagrees.

Mr Waugh submits that it is apparent that this approach proceeds from the same flawed premise that it was the belief that it was the unmetabolised terfenadine that was the sole cause of the TdP. He goes on to say that, in practice, if the skilled person were to have thought that this test would indicate which compounds were liable to cause extended QT intervals/TdP then it is more likely that they would first test astemizole and it would be seen that this drug too blocked potassium channel. A similar test carried out on cetirizine and loratadine would show that these did not have a significant blocking effect on the potassium channel. The skilled person would not consider that in the circumstances a drug that was as similar to terfenadine as the acid metabolite would be likely to show an effect different to either terfenadine or astemizole, which had much greater structural differences. A conclusion to jump to synthesising the metabolite and testing it in the way described in Chen is manifestly driven by hindsight.

I reject that submission. First, I do not think that the premise is correct if by “belief” Mr Waugh intends to indicate certainty. The test is being undertaken precisely to eliminate the lingering doubt that the acid metabolite does have a part to play in causing the adverse cardiac effects. If “belief” is used to contrast a lingering doubt with a real alternative hypothesis, then from what I have already said, the pointers do appear, by 1992, to have been in favour of the Monahan/Mathews hypothesis. In addition, it seems odd to me that you would first test other drugs (eg astemizole) when what you ultimately want to know is whether the acid metabolite can be developed as an effective drug in its own right. I found Professor Frew’s rejection of this suggestion convincing.

Against the conclusions that it was obvious to test the acid metabolite in the way suggested by Professor Frew and that the (satisfactory) result of that test would have resulted in the development of the acid metabolite as an antihistaminic drug in its own right, two points were put. First, Doctor Morley had an alternative route; secondly, it is said that the FDA materials pointed to the conclusion that the acid metabolite might have a part to play in the adverse cardiac effects.

As to Dr Morley’s alternative route, he raised the possibility that the skilled team would reject developing the metabolite but would instead attempt to synthesise a new antihistamine which did not have the QT/TdP problem, perhaps based on the old sedating antihistamines such as ketotifen. For Dr Morley, the more obvious course was to take an antihistamine which did not have the problem which could be used as a basis for getting a new and better antihistamine; the team would try its hardest to knock out the sedation. He considered that that would be an easier thing to do than to try to eliminate the cardiac effect which is due to material entering the cell and interacting with the ion channels.

It seems to me, however, that this is beside the point. The question is not whether Dr Morley’s alternative route was a preferable, or more obvious, route but whether it would have been an obvious course to subject the acid metabolite to the I_k test and, after obtaining the satisfactory result, to move on to develop it. That a positive outcome of the test would have been obtained is shown by what actually happened: see example 3 of 443 and 860. Moreover, Dr Morley accepted that it was at least a strong possibility that his preferred test – the Langendorff heart test – would also have given the same result.

Further, it seems to me that Dr Morley’s approach ignores the very good safety record of terfenadine itself. His concerns were largely based on the perceived risk that the acid metabolite might have. But there would be no reason for thinking that there were greater risks in respect of the acid metabolite than in respect of terfenadine, a drug which remained on sale until the acid metabolite was in fact marketed as a drug in its own right some years later. This was not a case of a parent compound which it had not been possible to bring to the market because of its associated risks. One might also reflect on the problems surrounding the development of the new drug advocated by Dr Morley bearing in mind that the development of the second-generation antihistamines must have been the result of attempts to do precisely what Dr Morley suggests – that is to say to produce an antihistamine without sedating side-effects.

Mr Waugh submits that there was no reason to develop the acid metabolite at all. To the contrary, it was only about 30% as potent as the parent drug, it was associated with the cardiac side effects seen and it was structurally very similar to terfenadine. There were thus no advantages only disadvantages. It is not, I think, with the benefit of hindsight that I agree with Professor Frew and Mr Watson that these factors, in the light of what was known in 1992, would not have deterred the skilled team from proceeding to test the acid metabolite as a start along the path to bringing to the market a drug which had all the advantage of terfenadine without its disadvantage in relation to adverse cardiac effects.

Mr Watson also makes the point, in relation to development, that it might well be quicker and cheaper to develop the acid metabolite rather than a new compound. This is because one knows that the acid metabolite has been made in man. Professor Frew considered that this meant that

“a lot of the toxicology issues which kill off drugs in early development have kind of been dealt with because patients have had quite significant quantities of this, the sort of level that you need for antihistaminic effects, without having any tangible level of side-effects up until this point. So that is why it is cheaper to go with this drug because you are not having to develop several hundred compounds and narrowing it down to one that you then put through clinical trials.”

I have already devoted some space to the transcript of the FDA committee meeting on 11 June 1990. Considerable reliance is placed by Mr Waugh on the materials relating to that meeting, in particular the reports of cases of adverse cardiac effect in the absence of measurable levels of serum terfenadine.

The matter brought to the committee was terfenadine’s ability to cause prolongation of the QT interval and TdP. The debate reflected the seriousness (or absence of it) of that issue and revolved around a relabelling and warning: this was not a case where a drug might need to be removed or have its product licence cancelled. During the course of that debate, many ideas came out. Those included a possible class effect (discussed elsewhere in this judgment). But it also included Professor Woosley’s proposition that it was the parent terfenadine which was responsible for the adverse cardiac effects. That being so, Mr Watson asks rhetorically how it can be inventive to check whether Professor Woosley’s view is correct. Ultimately, the answer to that question in my view is, as will be seen, that it cannot be.

In relation to the data contained in the paper submitted by Merrell Dow to the FDA committee (see paragraphs 79ffabove) Mr Watson’s submissions are these:

Effectively, Mr Watson says that, even if he had to rest with the FDA material which he says he does not, those cases would not be enough to deter the skilled team from continuing with a test to check Professor Woosley’s proposition.

Now, these may represent perfectly sound, and possibly correct, explanations of these cases which do not fit with the science as it is now known. But I agree with Mr Waugh when he says, in effect, that these explanations are reached with the benefit of hindsight and that, taken as part of the picture in 1990 they would cast doubt on the correctness of the Monahan/Mathews hypothesis. They would not destroy it precisely because there might be an explanation for the results. But they would weaken it and lead one to question whether it was obvious to test the metabolite. Moreover, Professor Frew may well have taken away from Garteiz and Sorkin more than was justified in terms of the extent to which the activity of terfenadine was attributable to the acid metabolite.

However, Mr Watson does not have to rest with the 1990 data. In particular, the skilled team would have Monahan (ie the paper and not just the case report) and Mathews. It might be said that Monahan was only one piece of information: but it was a piece of published information and, significantly, is confirmed by Mathews representing the view of Merrell Dow itself. The skilled team would also have Chen. Whatever Mr Waugh’s objections to the matters which I refer to in the next following paragraph, Monahan and Mathews have always formed part of Teva’s case, and Chen was added by amendment during the course of proceedings.

The skilled team would also have had information about the 25 cases referred in Woosley I: as to this information, see paragraph 125 above. Mr Waugh complains that this matter had not formed part of Teva’s pleadings or of its written case, nor was it dealt with in Professor Frew’s evidence. It is true that it was not identified expressly by Teva prior to the hearing, but given that it is part of the material that the skilled team would have, I do not think that I should ignore it. That is so for at least two reasons. The first is that it was only at a very late stage that the Defendants introduced the transcript of the FDA hearing and the Merrell Dow paper which went to the committee. In the absence of that material, the pointers were all towards the Monahan and Mathews hypothesis. There can be no objection to that material being met by something further which shows that it is to be treated with very great circumspection. Indeed, Mr Watson says that it is very unsatisfactory that Mr Waugh chose to cross-examine Professor Frew on the cases considered by the FDA committee but did not mention the later cases now relied on . Secondly, I do not see that the Defendants have been prejudiced by the introduction of this material. It must have been within the knowledge of Merrell Dow who could have instructed Dr Morley in relation to it. In any event, Dr Morley did not suggest he was not able properly to deal with it.

For his part, Dr Morley was further concerned with the uncertainty about whether, even knowing the results of a test such as the cat myocytes test, one could bring the acid metabolite to market. He pointed out that one would still have to do clinical trials and indeed that one would not know whether the acid metabolite actually caused TdP until 50 -100,000 patients had used it. Asked why he remained worried about the metabolite, he replied

“A. Because you take groups of evidence and you are making the assumption that this test is predictive of torsades de pointes and I would put it to you that not until several years had lapsed could you draw that conclusion. That is why I made reference to a review of the issue of how you detect torsades de pointes in patient populations and the conclusion is that you cannot rely on any pre-clinical test to make this conclusion.

Q. No. The ultimate safety of course depends on wide exposure to the population, but decisions to carry out research have to be based on things like in vitro tests of the ability to Ik channel block.

A.

Yes, that is correct.”

Mr Watson says that the Defendants cannot say, to defeat the obviousness attack, that one would not know whether it would be possible to get the acid metabolite to market. The Patents do not show that either and indeed the acid metabolite was not marketed until 1996. When assessing obviousness, the question is simply whether it would be obvious to subject the acid metabolite to the kind of in vitro assays, such as the Chen cat myocytes test, and beyond that it is not necessary to go. That, at the end of the day, is all that the Sepracor patents themselves contain and the Merrell patent does not even contain that.

In the light of all the material which I have seen and taking account of the argument addressed above, the position, in my judgment, is as follows:

My conclusion therefore is that there was nothing, or at least nothing sufficiently significant, to prevent the testing of the acid metabolite as being something worth trying with a good prospect of success. And once the last piece of the jigsaw had been put in place as a result of positive outcome of the I_k channel test, there was nothing to make it other than obvious to develop the acid metabolite to make a medicament for producing an antihistaminic effect or the treatment of allergic rhinitis and urticaria in people (whether hepatically impaired or not). Accordingly, Teva succeeds in its case on obviousness on that ground.

In addition, it is said that the Patents are obvious over Carr. I have dealt at length with Carr in the context of novelty. The material which I have covered in that context is also relevant to obviousness. Mr Watson explains that Carr teaches using the acid metabolite as an anti-histamine. The public are permitted, he says, to adopt this teaching with obvious modifications to it: it would be wrong to prevent people from doing something which is merely an obvious extension to what has been available before. In the present case, so the argument runs, there is not even a modification; instead, there is simply an application of the teaching to sub-groups of those patients envisaged by Carr (eg tohepatically impaired patients suffering from allergic rhinitis).

Mr Waugh says that Carr provides no basis whatsoever for saying that the use of the acid metabolite as the active agent in a medicament would avoid the cardiac side-effects seen on administration of terfenadine. He says that the Claimants’ case on Carr can be no stronger than over cgk; in fact it is weaker as, by 1992, the skilled reader would know that the acid metabolite was one of a number of compounds which had been available to Merrell Dow yet there had been no development of this compound whether as a bronchodilator or otherwise.

Mr Waugh is, of course, correct in saying that Carr provides no basis for saying that the use of the acid metabolite as the active agent in a medicament would avoid the cardiac side-effects seen on administration of terfenadine. That is hardly surprising since the side-effects of terfenadine were not known at the time. But that is not an answer to the point that Carr teaches the use of the acid metabolite as an anti-histamine which, for the reasons already given, I think it does. That it avoids the side-effects is, in my view, a discovery and not an invention.

Further, for the reasons given in relation to second medical use, I considered that use for the treatment of allergic rhinitis and urticaria (whether in hepatically impaired patients or not) would not be a new therapy in Bristol-Myers terms. But if that is wrong, I find it hard to see how it is not obvious to use the acid metabolite in the treatment of allergic rhinitis and urticaria (whether in hepatically impaired patients or not) once it is accepted that Carr teaches the use of the acid metabolite as an anti-histamine without qualification.

Accordingly, I consider that Teva is entitled to succeed on this ground too.

That leaves claim 12 of 443 which, as I have said, Mr Watson accepts has bare novelty because Carr discloses 50 mg compositions whereas this claim discloses a 60 mg composition. As to that, Professor Frew says this in his first report (and was not challenged on it):

“Terfenadine was marketed at a dose of 60 mg twice daily or 120mg once daily. Given that almost all of the antihistaminic activity seen after an oral dose of terfenadine seemed to be due to the acid metabolite, it would be expected that an effective dose of the acid metabolite would be 60mg/day or above. This is based on knowing that approximately 50% of an orally administered dose of terfenadine ends up being excreted as the acid metabolite (suggesting that for every 10mg of terfenadine ingested, about 5mg of acid metabolite is produced). Hence if the acid metabolite was fully absorbed and available, an oral dose of 60mg of the acid metabolite should have equivalent antihistaminic activity to an oral dose of 120mg of terfenadine. Clearly it would be necessary to conduct standard dose-finding studies to determine the actual minimum effective dose and the optimum dosing frequency (i.e. once daily, twice daily etc).”

I accept that evidence. It leads to the conclusion that the dosage specified in the claim was obvious. Mr Waugh comments that Teva’s case on obviousness has not been presented along Windsurfing lines. Whilst acknowledging that that approach is not compulsory, he says the failure to follow it can lead to error. Jacob LJ’s reformulation of the approach In Pozzoli SpA is as follows:

As it happens, I think that all of these matters are addressed in Mr Watson’s argument albeit without a rigorous adherence or formulaic approach. Thus step (1) has clearly been dealt with.

As to steps (2) and (3), there could be debate about whether the inventive concept was (i) using the acid metabolite as an anti-histaminic medicament or (ii) using it as such medicament whilst avoiding cardiac side-effect or (iii) more narrowly, in the case of each of (i) and (ii), to treat allergic rhinitis and urticaria. However, as Jacob LJ said, if a disagreement about the inventive concept of a claim starts getting too involved, the sensible way to proceed is to forget it and simply to work on the features of the claim. I think that all of these aspects have been fully addressed by Mr Watson. And step (4) has clearly been thoroughly debated. Accordingly, I do not think that the approach which has not followed Windsurfing to the letter has led Mr Watson (or I hope myself) into error.

Claims 1, 5 and 12 of 976, claim 1 of 443 and claims 1 and 3 of 860 are invalid as lacking novelty. Those claims as well as claim 12 of 443 are invalid for lack of inventive step. Teva is entitled to declarations accordingly.