Baxter Healthcare Corporation & Anor v Abbott Laboratories & Anor

This is an action in which the claimants (“Baxter”) seek a declaration of non-infringement in respect of EP (UK) 0967975 (“the Patent”), and revocation of the Patent.

The Patent is concerned with the degradation of a particular fluoroether in the presence of Lewis acids. Before turning in detail to the disclosure of the Patent, it would be helpful to indicate a little about fluoroethers and Lewis acids. It is tempting also to sketch out the history of the invention, but this is a temptation that must be resisted. When I come to interpret the patent specification and attempt to set out what it discloses, there are a number of obscurities and difficulties which become understandable when the history of the invention is known, but that is not material admissible on the question of interpretation, which is an objective assessment of the document itself. In the case of a patent, one is interested in the result, not how one got there.

The fluoroether compositions with which the Patent is concerned are shown in the general formula in [0018]. The sevofluranes according to this general formula are anaesthetic compounds, examples of which are sevoflurane, enflurane, isoflurane, methoxyflurane and desflurane. Each is said to be potentially the subject of degradation caused by Lewis acids, and the inventive concept lies in including in a fluoroether composition a compound that inhibits Lewis acids, referred to in the specification as a “Lewis acid inhibitor”. The greater part of the description (the examples aside) is couched in terms of fluoroether compounds generally, but the claims are limited to sevoflurane.

It is necessary to explain only a little about Lewis acids. Generally speaking, what one recognises as an acid is a material which on dissolution in water gives a proton (i.e. a hydrogen ion, in fact appearing as the hydronium ion H₃O⁺) and an anion such as the sulphate, nitrate or chloride ion. The classic acid-base reaction involves the reaction of these species with, for example, a metal oxide, yielding water via the reaction between the hydrogen ion and the oxide group. Materials yielding a proton in aqueous solution are called Brønsted-Lowry acids. Lewis acids are a generalisation of such materials born of Lewis’s insight into the nature of the chemical bond. A Lewis acid is defined as a material having two available electron sites in the outermost shell of electrons capable of accepting a pair of electrons from another species or compound. A Lewis base (a synonym for a “Lewis acid inhibitor”) is a species or compound which has a pair of electrons which are available for bonding with an electron-deficient species to form a covalently bonded compound. When one talks of electron-deficient species, one is not talking about electron-deficiency in terms of charge – the compounds in question are electrically neutral – but about the occupancy of electron orbitals or shells. For present purposes, it is only necessary to be aware that water is a Lewis base and that it is the only Lewis base with which the case is concerned. Apart from the fact that one would expect a graduate chemist to be aware of the nature of Lewis acids and bases, it is not necessary to consider this aspect of the chemistry in any more detail.

The Patent starts with a description of the field to which the invention relates and after a brief description of fluoroether compounds in [0002] states the problem with which it is concerned in [0003]:

“Although fluoroethers are excellent anaesthetic agents, it has been discovered that some fluoroethers experience stability problems. More specifically, it has been determined that certain fluoroethers, in the presence of one or more Lewis acids, degrade into several products including potentially toxic chemicals such as hydrofluoric acid. Hydrofluoric acid is toxic by ingestion and inhalation and is highly corrosive to skin and mucous membranes. Thereupon, the degradation of fluoroethers to chemicals such as hydrofluoric acid is of great concern to the medical community.”

The manner in which the degradation takes place is described in [0004] and [0005]:

“[0004] Degradation of fluoroethers has been found to occur in glass containers. The degradation of fluoroethers in glass containers is believed to be activated by trace amounts of Lewis acids present in the container. The source of the Lewis acids can be aluminium oxides, which are a natural component of glass. When the glass wall becomes altered or etched in some manner, the aluminium oxide become exposed and come into contact with the contents of the container. The Lewis acids then attack the fluoroether and degrade it.

[0005] For example, when the fluoroether sevoflurane is contacted with one or more Lewis acids in a glass container under anhydrous conditions, the Lewis acid initiates the degradation of sevoflurane to hydrofluoric acid and several degradation products. The degradation products of sevoflurane are hexafluoroisopropyl alcohol, methyleneglycol bishexafluoroisopropyl ether, dimethyleneglycol bishexafluoroisopropyl ether and methyleneglycol fluoromethyl hexafluoroisopropyl ether. The hydrofluoric acid proceeds to further attack the glass surface and expose more of the Lewis acid on the glass surface. This results in further degradation of sevoflurane.”

After setting out formulae for the various degradation products so identified, the invention is summarised in [0008] and [0009]:

“[0008] The present invention involves a stable anaesthetic composition that contains a fluoroether compound having an alpha fluoroether moiety having added thereto an effective stabilizing amount of Lewis acid inhibitor. The fluoroether compound is sevoflurane and the Lewis acid inhibitor is water. No soda lime is present in the composition. The composition can be prepared by adding the Lewis acid inhibitor to the fluoroether compound, by adding the fluoroether to the Lewis acid inhibitor, or by washing a container with the Lewis acid inhibitor and then adding the fluoroether compound.

[0009] The present invention also involves a method for stabilizing a fluoroether compound having an alpha fluoroether moiety. The method involves adding an effective stabilizing amount of a Lewis acid inhibitor to the fluoroether compound to prevent the degradation of the fluoroether compound by a Lewis acid. The fluoroether compound is sevoflurane and the preferred Lewis acid inhibitor is water.”

It is convenient at this point to set out the claims as they presently appear so far as relevant. Claim 1 is as follows:

“An anaesthetic composition comprising:

a quantity of sevoflurane; and

a quantity of water providing a concentration of water in said anaesthetic composition of between 0.015% w/w and a saturation level of water in said quantity of sevoflurane, wherein no soda lime is present in the composition.”

Claim 1 is the only claim to a composition. The remaining claims are all method claims and are all directly or indirectly dependent upon Claim 2, which is as follows:

“A method of preventing degradation by Lewis acid of a quantity of sevoflurane, the method comprising the steps of:

providing a quantity of sevoflurane;

providing a Lewis acid inhibitor in an amount sufficient to prevent degradation of said quantity of sevoflurane by a Lewis acid; and

combining said quantity of sevoflurane and said Lewis acid inhibitor.”

Claim 3 claims a class of Lewis acid inhibitors, one of which is water. Claim 4 is a specific claim to the method according to Claim 2 in which the Lewis acid inhibitor is water, and Claim 5 claims a range of concentrations of water in a method according to Claim 4 from about 0.0150% w/w to about 0.1400% w/w in the resulting formulation.

The evidence was that 0.1400% w/w (1400 ppm) represents a saturated solution of water in sevoflurane. The lower limit of 0.0150% w/w for water is explained further in [0026]:

“The method of the present invention uses an effective stabilizing amount of a Lewis acid inhibitor. It is believed that the effective stabilizing amount of Lewis acid inhibitor that can be used in the composition is about 0.0150% w/w (water equivalent) to about the saturation level of the Lewis acid inhibitor in the fluoroether compound. As used herein, the term “saturation level” means the maximum solubility level of the Lewis acid inhibitor in the fluoroether compound. It will be appreciated that the saturation level may be temperature dependent. The saturation level also will depend on the particular fluoroether compound and the particular Lewis acid inhibitor being used in the composition. For example, when the fluoroether compound is sevoflurane and the Lewis acid inhibitor is water, the amount of water employed to stabilize the composition is believed to be from about 0.0150% w/w to about 0.14% w/w (saturation level). It should be noted, however, that once the composition is exposed to Lewis acids, the amount of Lewis acid inhibitor in the composition may decrease as the Lewis acid inhibitor reacts with the Lewis acid to prevent the unwanted degradative reaction of the Lewis acid inhibitor with the composition.”

This is in contrast to the “starting” anhydrous fluoroether compound which is discussed in [0018] as follows:

“The anaesthetic composition of the present invention contains an anhydrous fluoroether compound. The term “anhydrous” as used herein means that the fluoroether compound contains less than about 50 ppm of water. [i.e. 0.0050% w/w] … ”

After briefly explaining the mechanism whereby the Lewis acid inhibitor forms a covalent bond with the acid and so prevents its reacting with the alpha fluoroether moiety of the fluoroether and causing degradation, the specification suggests various manners in which the composition can be prepared. The list, in [0029], includes suggestions to either wash or rinse the container with Lewis acid inhibitor and then fill it with the fluoroether compound. A partial drying step after washing or rinsing is said to be optional, and sealing is said to follow upon the addition of the fluoroether. Alternatively ([0030]) the Lewis acid inhibitor is added to a dried container prior to filling with the fluoroether compound, or the Lewis acid inhibitor is added directly to the container already containing the fluoroether compound. The third possibility ([0031]) is to fill the container with the fluoroether compound under “humid conditions”, allowing the filled compound to stand in a humidity chamber for a sufficient amount of time to allow the water to accumulate in the container. Finally, [0032] observes that the Lewis acid inhibitor can be added to the composition at any appropriate point in the manufacturing process, e.g. at the final manufacturing step before filling into shipping containers, e.g. a 500-litre shipping container. It is suggested that appropriate quantities of the composition can be dispensed from such a container and packaged in containers of more suitable size for use, and it is also suggested that small quantities of composition containing appropriate amounts of Lewis acid inhibitor can be used to wash or rinse containers.

Sevoflurane is very hygroscopic, and unless special precautions are taken it will absorb water from the atmosphere. Dr Lessor, Baxter’s principal expert witness, so described it and was not challenged.

The specification then turns to the examples. It is convenient to consider the examples as their significance becomes clearer. There is no doubt that they are very significant, because they provide the only illustration of a system throwing light upon important features of the method claims, to which I shall return.

The patent in suit is obviously directed towards a manufacturer of sevoflurane – or of fluoroether compounds generally – who is interested in improving the stability of the composition which he sells. A manufacturer of sevoflurane will be concerned primarily in the manufacture of sevoflurane for anaesthetic purposes, and I do not believe that it was ever suggested that there was an extensive interest in these materials other than for use in production of anaesthesia. Professor Kilburn, one of Abbott’s expert witnesses, took the view that the Patent was addressed to people interested generally in the stability of alpha fluoroether ethers. Whether the skilled person is employed by a manufacturer of sevoflurane or by a manufacturer of other alpha fluoroethers does not seem to me to be of particular concern: the only compound which matters to the claims is sevoflurane, and when considering the validity of those claims and the nature of the disclosure, it is to the chemist employed by the sevoflurane manufacturer that it is necessary to refer.

It is agreed that such a chemist will have as part of his common general knowledge the contents of the primer which is to be found in bundle 2 at tab 11. I shall refer to this document from time to time, as necessary. The issues of construction are threefold. So far as Claim 1 is concerned, the question is what is meant by “an anaesthetic composition”. This question arises because certain alleged prior disclosures may or may not be relevant to the validity of the claim, depending upon whether they are or are not “anaesthetic compositions”. The second issue arises in relation to the words of the claim that reflect the heart of the invention. Claim 2 refers to “a method of preventing degradation” and to incorporating a Lewis acid inhibitor in “an amount sufficient to prevent degradation” of the quantity of sevoflurane. What is meant by “preventing degradation”? Finally, the last feature of Claim 2 refers to “combining” the quantity of sevoflurane and the Lewis acid inhibitor. What here is meant by “combining”? As is so often the case, the full significance of the points that arise in relation to construction can be appreciated only in the context in which they arise. At this point, I shall content myself with setting out merely a statement of the applicable principles. The task of construction is to determine objectively the meaning of the words used by the patentee to describe and claim his invention in their context in the specification. The statement of the law in Kirin Amgen [2005] RPC 9 (page 169) exhaustively discusses and describes the process of construction of a patent. The approach to construction there described results in a construction of the document which accords with that mandated by Art. 69 of the European Patent Convention, which describes the basic dichotomy between fair protection for the patentee on the one hand and a reasonable degree of protection for third parties on the other. As Lord Hoffmann’s speech – [48] and [52] in particular – suggests, the uncertainty necessarily arising in marginal cases can frequently be alleviated by the “structured” approach represented by the so-called Protocol questions set out in Improver Corporation v. Remington Consumer Products Ltd [1990] FSR 181, 189. These questions are guidelines and can be useful. To approach the question of interpretation, whether by employment of the guidelines or by approaching the purposive interpretation of the words in their context without the assistance that the guidelines provide, necessarily requires the question to be properly formulated. It is accordingly necessary now to consider the context in which the problems of construction that I have identified above arise.

Having given this introduction to the issues in the case, I must say a word about the expert witnesses. I have already referred to Dr Lessor, who was Baxter’s principal witness. Baxter also called Dr Woodhouse, who had extensive experience in pharmaceutical development. Abbott called Professor Chambers, a distinguished expert in organofluorine chemistry, and Professor Kilburn. The opposing party asked me to treat the evidence of, in particular, Professor Kilburn and Dr Lessor with caution.

This action is one branch of a wide dispute concerning this patent and its equivalents and derivatives round the world which is occupying Baxter and Abbott. As I understand it, Dr Lessor, who has been involved with the development of sevoflurane by Baxter and its predecessor in this field, Ohmeda, since 1997, has already had a substantial involvement in the corresponding proceedings in the US. I do not think that this affected his evidence in an objectionable manner, although I have little doubt that he was already familiar with some of the criticisms of Baxter’s case which were made both in Abbott’s expert reports and in cross-examination.

A rather more serious attack was made on the quality of Professor Kilburn’s evidence. Professor Kilburn is the Head of the School of Chemistry at Southampton University and is presently the Vice-President of the Organic Division of the Royal Society of Chemistry. Mr Floyd QC criticised him for possessing a degree of enthusiasm for Abbott’s case, which led him to attempt to correct an answer that he had given in cross-examination after considering the transcript overnight. The problem was infringement. Professor Kilburn wished to say that there was enough water in Baxter’s product that, in combination with the resin lining of the bottle, the function requirements of Claim 2 were met. He said that the amount of water (< 130 ppm) was of itself sufficient. This leads to potential problems with validity on the basis of the Invention Disclosure Statement, of which more below. Professor Kilburn sought to withdraw the answer he had given, and Mr Floyd submits that this indicated excessive enthusiasm for Abbott’s case – see transcript pages 722-5, 744. At the same time, Mr Floyd is careful not in any way to impugn the Professor’s integrity. The complaint is that from time to time excessive enthusiasm took the edge off the high degree of objectivity expected of expert witnesses. I think there may be something in this, but ultimately I do not think it matters. After all, as Mr Floyd accepts, the Professor did give answers that clearly favoured Baxter’s case. While accepting that a slight reservation might be appropriate, therefore, I am not conscious that it has affected the view I take of the issues as a whole.

In a confidential Product and Process Description (“the PPD”) Baxter set out in purported compliance with the requirements of s.71 of the Patents Act 1977 a description in writing of the product which they propose to import into the United Kingdom and the method by which it is prepared. The PPD was amended on a number of occasions, so as to contain references to certain experimental work conducted by Baxter and set out in various documents (Annexes A – E and DX 322), each of which augments the list of identifying criteria by which the material forming the subject matter of the action for a declaration of non-infringement is to be characterised. The successive amendments to the PPD to some extent reflect the lack of any positive case of infringement advanced by Abbott. Abbott’s policy throughout these proceedings has been to decline to make any positive case of infringement, but to point out what it perceives to be deficiencies in the PPD, deficiencies which it alleges are sufficient to leave the question “infringement or no” uncertain. Thus, to take a particular example, Baxter describe the container in which they propose to package their Sevoflurane. This package is lined with a particular defined epoxyphenolic resin. While not alleging that the epoxyphenolic resin of the Baxter container does in fact act as a Lewis acid inhibitor, nevertheless Abbott suggest that it is for Baxter to perform the necessary experimental work:

“Given the theoretical grounds for expecting the liner to have Lewis acid inhibition properties, and given the inhibition that is shown with the 10 mg sample in the hypothetical data in Annex E, the evidential burden certainly lies with Baxter to show that the liner is not acting as an inhibitor across the PPD.”

Bizarrely, I conducted the entire trial not knowing whether the liner material – however theoretically capable of acting as a Lewis acid inhibitor – did in fact do so.

Other potential sources of Lewis acid inhibitor, in particular an amine which is a potential by-product of the synthetic route to sevoflurane adopted by Baxter, were contended by Abbott to be a potential Lewis acid inhibitor, but by characterising Baxter’s sevoflurane as containing less than 0.5 ppm of two specified amines, Abbott accepted that no significant Lewis acid inhibition would take place.

The next characterising features to which I should refer are that numbered 2 on page 2 of the document and those numbered 5 and 6 on page 3. Together, these read as follows:

“2.

the product in its containers will degrade after 24 hours at 55 degrees Centigrade (as indicated by total impurity levels measured by gas chromatography of greater than 300ppm and pH values less than 4.0) when challenged with the amounts and type of Lewis acid set out in … [Annex C];

…

5.

the product in its containers will degrade substantially as shown in the attached graph of total impurities against time (Annex E) at 55 degrees Centigrade (as indicated by total impurity levels measured by gas chromatography of greater than 300ppm) when challenged with 10, 25 and 50 mg of the type of Lewis acid set out in the attached Pharmaceutical Development Report … [DX 322], Study A, …

6.

the same product in the same containers held under the same conditions as set out in 5 above but without such added Lewis acid will remain below 300 ppm total impurities over the same period.”

These paragraphs caused considerable confusion at the trial: DX 322 recorded an experiment done for the purposes of litigation, which is, of course, inadmissible as evidence of the results said to be proved thereby, unless the consent of the court is given or the standard directions for adducing experimental evidence (which were made in this case) are complied with. I ruled that the experiment could not be relied on in this way. But this is not the function of these experiments as they are used for the purpose of the PPD. So far as the PPD is concerned, the question can be framed like this: if experiment DX322 is performed, and if it has the result that is recorded, does that show that the sevoflurane either does infringe or does not infringe, or would that result, if obtained, be equivocal?

Interestingly, paragraph 2 of the PPD criteria set out above is criticised by Abbott on the grounds that it is conclusory. They say that mere use of the term “will degrade” is just a conclusion, and the statement “to be of any meaning has to be read in conjunction with the parenthesis”, i.e. as indicated by total impurity levels measured …. Further, Abbott continue, it throws up issues of what Baxter means in this document by “degradation”. It is not unfair, I think, to observe that this is indeed the question which arises on the claims and, with that introduction, I proceed to the question of construction which I have described above.

One other point needs to be borne in mind. For this I use Abbott’s words:

“It is common ground that degradation can never be prevented absolutely.”

It will be apparent from the discussion that I have set out above that the meaning of this word has given rise to considerable difficulty. I have already discussed the general disclosure of the patent specification, and I now turn to the examples which are said to be examples of the invention. Example 1 is concerned, as are the other examples, with glass bottles, which are made of so-called type III glass, which consists mainly of silicon dioxide, calcium oxide, sodium oxide and aluminium oxide. The last of these materials, aluminium oxide, is known to be a Lewis acid. While glass of this type does not, according to the specification, normally act as a Lewis acid, it may do so if the surface is attacked, exposing active Lewis acid sites. What is done for the purposes of the examples (see [0035]) is to add various amounts of activated alumina to 20 ml of sevoflurane. There are nine combinations of conditions: column 1 represents anhydrous sevoflurane containing 20 ppm water to which no additional water is added; column 2 represents sevoflurane containing 100 ppm water which has been added (“spiked”) to the sevoflurane; and column 3 a sample containing 260 ppm water. Three doses of activated aluminium oxide are used: 50 mg, 20 mg and 10 mg. The deliberate addition of a Lewis acid to sevoflurane in this way was referred to at the trial as an “insult” or “challenge”. As can be seen, the question is, what effect the addition of water has. Does it prevent the sevoflurane from degrading or does it not?

[0036] gives the answer, and I quote:

“It will be appreciated that 20 ppm Water is equivalent to 0.0022% w/w Water. The samples were placed at 60° C and analyzed by gas chromatography after 22 hours. Figure 1 shows that in the presence of the same amount of aluminium oxide (50 mg) that the degradation of sevoflurane decreases with increasing amounts of water (Row A from Table 1). A similar trend was observed for 20 mg and 10 mg of aluminium oxide (Rows B and C)”

Figure 1 does not set out the results for the experiments for Rows B and C. It is concerned only with Row A, i.e. 50 mg alumina. It will be seen that the picture for 260 ppm water is very different from that for 100 ppm water, in that there is some suggestion that the impurity spikes (labelled S1, P1, HFIP and P2) are somewhat diminished in height, but of course this is all rather qualitative. However, it is plain that at the 50 mg level of insult, there is substantially more degradation of the sevoflurane at water contents of 20 ppm and 100 ppm than there is at 260 ppm. So there the skilled person perhaps discerns the invention in action, a level of 260 ppm water being sufficient to produce a real improvement. Obviously, there is nothing in the text of the Patent suggesting that an insult equivalent to 50 mg of activated alumina in 20 ml sevoflurane is in any way characteristic of the sort of insults that would be obtained in practice. This is perhaps unsurprising, because it was agreed on all sides that such insults are exceptionally unusual and the only two insults of which I have any usable evidence at all are where the Lewis acid in question was provided by a corroded valve on one of the 500-litre bulk containers of Abbott’s sevoflurane to which I have referred, and also a recent event in which it appears that a particular design of vaporiser, part of the anaesthetist’s equipment, caused degradation of some of Baxter’s sevoflurane. The former event gave rise to the patent in suit and the latter is still under investigation. I find, therefore, that there is no material in the case or elsewhere from which any conclusion can be drawn relating to the likely magnitude of any Lewis acid insult to a particular quantity of sevoflurane during either manufacture, storage, transport or administration. On the other hand, everybody appears to have been agreed that the sort of exposure one would expect would be less than the equivalent of 50 mg activated aluminium oxide in 20 ml sevoflurane. Further than that, it is not really easy to go.

Although the question of the magnitude of the insult one can expect in practice may seem to be a point on the edge of the claim, in reality I do not think that it is. Claim 2 (but not Claim 1) contains the functional restriction that the Lewis acid inhibitor should be present in an amount sufficient to prevent degradation of the stated quantity of sevoflurane by a Lewis acid, but does not provide any criterion by which this amount is to be computed. But degradation of sevoflurane is not inevitable. Much sevoflurane has been sold over many years – the quantities are in the tens or possibly hundreds of tons – but so far as the evidence went, there have only ever been the two incidents of degradation by Lewis acid that I have described above. So the question, what is “sufficient to prevent degradation”, becomes extremely acute. The problem is caused entirely by a failure to specify what I might call a standard insult. This is why a proper analysis of the disclosure of the examples becomes important. They provide the specification’s only explicit pointers in the direction of quantities and kinds of insult.

Example 2 is concerned not with activated aluminium oxide but with glass. Two heat-sealable ampoules made of type I glass are filled, one with 20 ml of anhydrous sevoflurane and the other with 20 ml of sevoflurane saturated with water. Both ampoules are flame-sealed, which I assume is the process that may activate sites on the glass surface, and subsequently autoclaved to accelerate the process of degradation. This experiment demonstrates that the saturation with water inhibited any degradation that took place, and the gas chromatogram, which is said to show undegraded sevoflurane, forms Figure 3.

Example 3 is equally concerned with glass ampoules activated, one assumes, by the process of heat-sealing. The experiment is not entirely straightforward, but it should be noted that Set A and Set B, each of five ampoules, differ only in that Set A shows, in those bottles that have been heat-treated and subsequently autoclaved to accelerate degradation, that the inhibition of the degradation was, as the specification states at page 6 line 55, at least 595 ppm of water, sufficient to inhibit the degradation of sevoflurane.

Example 4 shows that in similar samples to those of Example 3 the amount of water required to inhibit degradation depends upon the conditions under which the sevoflurane is degraded. Thus, at 40° C for 200 hours, water levels higher than 206 ppm inhibit degradation, but for the samples stored at 60° C for 144 hours or longer, water levels higher than 303 ppm inhibit the degradation. The author of the specification indicates that the data suggest that as temperature increases, the amount of water required to inhibit the degradation of sevoflurane will increase.

Example 5 is again an example concerned with glass bottles, this time activated type III glass. These were bottles which had already been used to store degraded sevoflurane. As one of the degradation products is hydrogen fluoride, the bottles exhibited a significant amount of etching inside (hydrogen fluoride attacks glass, eventually dissolving it). The bottles used in the experiment were selected for significant amounts of etching and washed out with non-degraded fresh sevoflurane. They were then filled with anhydrous sevoflurane and heated at 50° C for 18 hours. This experiment showed a very substantial degradation over the 18 hours, and the interpretation placed upon the results is that the etched bottles had been “activated” by degraded sevoflurane, and that the activated glass surfaces thus served as initiators for the degradation of fresh sevoflurane. The experiment says nothing about inhibition of degradation.

Example 6 is again concerned with the degradation of sevoflurane in type III amber glass bottles, and the scheme of the experiment was to divide the bottles into two groups, a control group and a study group. The control group and the study group were both rinsed several times with anhydrous sevoflurane and then the control group were filled with 100 ml of anhydrous sevoflurane, while the study group were filled with 100 ml of sevoflurane containing about 400 ppm of water. Again, the bottles were heated at an elevated temperature, 50° C, for 18 hours, and the presence of two degradation products, those referred to as HFIP and P2, were assayed by gas chromatography. It is clear from these results that both the study group and the control group showed generally similar increases in HFIP, but that the study group did not show any substantial increase in P2. The results from both Experiments 5 and 6 are presented together in Figures 4 and 5, Figure 4 showing the results for P2 and Figure 5 the results for S1 (which is not set out in Table 6). The conclusion sought to be drawn from these two sets of results recorded in Tables 5 and 6 and in Figures 4 and 5 is that “the degradation of sevoflurane is inhibited by the addition of water at 400 ppm”.

Finally, Example 7 is again concerned with type III amber glass bottles that have been activated. The bottles used were five of those that had been used in Example 6, and the sevoflurane employed was water-saturated – i.e. 1400 ppm initially. The five bottles were placed on a roller and rolled for two hours “to allow the water to coat the activated glass surfaces”. The water-saturated sevoflurane was drained from each bottle and replaced by 100 ml of sevoflurane containing 400 ppm of water. Table 7 sets out the results for impurity content after 18 hours, 36 hours and 178 hours of heating at 50° C. The results are said to demonstrate that the degradation of sevoflurane was “greatly inhibited by treating the activated glass surface with water saturated-sevoflurane prior to heating”.

A number of observations may be made about these examples. The first is that there is no attempt to indicate what a typical insult might be. Second, all the examples, bar Example 1, are concerned with 250 ml glass bottles, and seem to provide some sort of “worst case scenario” in that the bottles in question had been etched by hydrofluoric acid or, in Examples 2, 3 and 4, activated by heating. Third, the word “inhibition” is plainly being used in a non-absolute sense, so that it makes sense to talk in [0048] of the degradation being greatly inhibited, in [0046] of the degradation being inhibited, and in [0045] of the degradation being significantly reduced.

[0026] describes the method of the present invention as using “an effective stabilizing amount” of Lewis acid inhibitor. Stability is of course the other side of the coin from degradation, but in the context of a pharmaceutical it is more straightforward to employ this concept rather than to worry about whether, in its context, inhibition of degradation means complete inhibition or some degree of inhibition less than absolute inhibition. In context, stability means: as stable as the marketing authorisation for the pharmaceutical in question requires. So, if the product in question must be stable for, say, two years at a stated temperature in order to satisfy its licence, then that is all that is required. If the composition satisfies such a requirement, together with the specific stability tests which will also be specified in the authorisations, I would use that to give a practical and sensible meaning to the words “an amount sufficient to prevent degradation”.

Unfortunately, the problem does not end there, because I do not believe that on a fair reading the specification provides any workable criterion for identifying a test that any manufacturer of sevoflurane can use to decide whether his quantity of water does or does not satisfy those words. He is told by the examples that, as one would expect, the amount of water required to inhibit degradation for a given insult increases with temperature (Example 4) and he is also told that the best results are obtained with 400 ppm water, a level which is said in [0027] to be the lower limit of a preferred range from 400 ppm to 800 ppm, a preferred range which finds no echo in the claims. I am attracted by the suggestion that [0026] should be employed so as to provide a numerical substitute for the idea of effective amount. The reason is that if one starts from the proposition that tons of sevoflurane have been sold without difficulty before the priority date, it would appear that in the ordinary run of business the material is not liable to degradation. An understandable reluctance to identify a typical insult means that the word “effective” is deprived of any substance because the specification gives no guidance as to what the inhibitor has to be effective to prevent, or, in the terms used in the claim, what the inhibitor must be sufficient to prevent. The claimants suggest that one should substitute for this criterion the criterion provided by [0026] of 150 ppm w/w (water equivalent) and upwards. This removes any doubt as to what the threshold value is. Furthermore, [0027] makes it clear that water is the preferred Lewis acid inhibitor, that its effective amount is believed to be from about 150 to about 1400 ppm, is preferably above 400-800 ppm, and that for any other Lewis acid inhibitor, a molar equivalent based upon moles of water should be used. Thus, 150 ppm represents a level which, on the teaching of the specification itself, is the lowest amount of water equivalent that the inventor believes to be effective.

Abbott submit that, in its context, the amount sufficient to prevent degradation is the amount that is sufficient to prevent unacceptable Lewis acid mediated degradation occurring, and, if this means that it is the amount sufficient to prevent the material going out of specification over its shelf-life, I think this is an acceptable construction. But it does not meet the difficulties which I have expressed above, because it still fails to provide any criterion by which prior art sevoflurane (which for present purposes must be assumed to be anhydrous and to satisfy its stability and impurity conditions in the licence) must be modified by the addition of water in order to comply with the claim. After all, if it did not degrade before, it is not going to degrade after.

The approach to this problem advanced by Abbott I found difficult to understand. Professor Kilburn and Professor Chambers both took the view that Table 6 of the Patent and the text associated with it in [0045] and [0046] indicated that Bottle 1 in particular, which contained a material with 669 ppm impurities, was undegraded. The total impurities in the study group were undoubtedly much lower than those in the control group in which water was not added, but [0045] contains the following statement:

“The amounts of degradants P2 (dimethyleneglycol bishexafluoroisopropyl ether) and S1 (methyleneglycol fluoromethyl hexafluoroisopropyl ether) were much less than those in Control Group 1 (20 ppm water). The HFIP concentration in the Study Sevo Group, however, was quite high and suggests that the glass surfaces were still somewhat active.”

The statement that the “glass surfaces were still somewhat active”, taken with the basic assumption of the Patent that it is the activation of the glass surfaces (accompanied, as it is, by the exposure of Lewis acid sites in the glass) that is responsible for the Lewis acid mediated degradation of the sevoflurane, makes it impossible, in my view, to suggest that the amount of water in this case was sufficient wholly to inhibit the Lewis acid mediated degradation of the sevoflurane. Furthermore, I do not find any suggestion that such a level of degradants in the accelerated stability test which this table in fact represents would be sufficient to satisfy a reasonable specification for sevoflurane intended for use with human beings. Table 7 seems to me to represent the desired result: within the limit of experimentation, there was no degradation after 178 hours at 50° C, a result achieved by first washing the activated glass bottles with water-saturated sevoflurane on a mechanical roller for approximately two hours “to allow the water to coat the activated glass surfaces”.

At the hearing Dr Lessor performed a calculation on the results for Bottle 1, which was the bottle in the study group that ended up with 669 ppm impurities. It is clear from the stated results that substantially all the impurities are HFIP, and Dr Lessor’s calculation simply assumes that there is a stoichiometric degradation of sevoflurane to HFIP and hydrogen fluoride and formaldehyde. Such a degradation will result in a material containing 400 ppm water with a pH of 2.3, pH here really being an indication of hydrogen fluoride concentration. There can be no doubt that this level of hydrogen fluoride represents unacceptable degradation. Professor Chambers had taken the view that the presence of HFIP in Example 6 was, in fact, an artefact of the scheme of experimentation, the HFIP produced in Example 5 being adsorbed onto the walls of the glass bottles, not being removed during the rinsing process with anhydrous sevoflurane described in [0043], and so representing the whole of the apparent degradation products in the study group. He told me that adsorbed HFIP at the beginning of the experiment would account for the whole of the impurity result at the end, and that the zero-hour reading which shows no HFIP present would be explained by the fact that all the HFIP present was in fact adsorbed onto the walls of the container and was not available to be measured. He thought of this possibility after having seen Dr Lessor’s calculation in X9. Thus, on a proper interpretation, the study group results in Table 6 are in fact the same as, or similar to, the 36-hour figures in Table 7.

I do not find in the specification, fairly read, any suggestion that the 669 ppm figure represents acceptable levels of degradation. On the contrary, the figures in Example 3 and Example 4 suggest a desire to get the total degradation products well down to something of the order of 100 ppm, as Baxter submit. I consider that the clear wording of [0045] indicates that the results for the study group in Table 6 are not sufficient. These are, in fact, the sorts of levels that sevoflurane shows, and a document put in evidence under Civil Evidence Act Notice by Baxter, consisting of two letters from the second defendant to Ms Delgado-Herrera, shows that the sevoflurane is stable under normal conditions and had been for six years prior to 1996. The typical total impurities of what is called Compound A and all other impurities has a bulk specification of 300 ppm w/w, but appears, as a practical matter, to have been held between 70 and 100 ppm.

I have come to the conclusion that unless Claim 2 is read in context as requiring that at least the lowest level of water which the specification says the patentee believes to be effective is present, that is to say 150 ppm, then the claim and the specification suffer from overwhelming difficulties of sufficiency of description. The basic problem is caused, I think, by the fact that the Patent specification nowhere specifies the degree of insult which the various containers used in Examples 2 to 7 have undergone. Without fairly brutal treatment, sevoflurane is stable according to its stated specifications, and accordingly, for the vast bulk of sevoflurane in commerce, the addition of water to any specified level has absolutely no effect. It is truly an addition intended to have a prophylactic effect against difficult-to-foresee but potentially damaging occurrences, of the kind in which the sevoflurane may contact Lewis acids. Given that the essentially unforeseeable nature of these occurrences prevents the skilled person from being able to decide upon a representative event a priori, the only reasonable approach is to substitute a hard criterion in the claim, supporting it by the patentee’s statement that that at least is the minimum. I take the view, for reasons which I shall set out below but have summarised, that unless this claim is treated as in effect coextensive with Claim 5, then Claims 2, 3 and 4 must be taken to define an invention which is not described either clearly enough or completely enough for it to be performed by a person skilled in the art, and are, accordingly, invalid.

This question arises because it is suggested, as I have indicated, that the epoxyphenolic resin liner of the aluminium container proposed to be used by Baxter is said to function, at least potentially, as a Lewis acid inhibitor and it is complained that there is no evidence in the case justifying me in drawing the conclusion that, in all circumstances falling within the PPD, it does not do so.

In this respect, Baxter have been fortunate. Recently, they had a problem with some vaporisers made by a particular manufacturer, in which it appears that their sevoflurane degraded over time. Of course, the lining of a tin can have no impact on what happens after the sevoflurane leaves the tin, which is the position with the vaporiser. But it throws incidentally an interesting light on the question of combination, since evidently the composition in the vaporiser is still the composition that was sold by Baxter; it degrades; and so the quantity of sevoflurane provided by Baxter is not, and cannot be, at this stage protected from degradation by any inhibiting effect of the liner. Accordingly, at this point, if the word “combining” is apt to include the provision of a liner in the aluminium potentially having Lewis-base effects, it has been uncombined by the time the sevoflurane is used. So what does combining mean in this context?

Unsurprisingly, Baxter submit that the whole thrust of the specification is towards something, preferably water, which is mixed with the sevoflurane. That is confirmed by the fact that the amount of inhibitor is defined, whenever it appears, upon a w/w basis, and it is further said to be physiologically acceptable – suggesting, at the lowest, that it is contemplated that the patient will inhale it.

Abbott, on the other hand, submit that “combining” would cover (for example) both the deliberate addition of additional water to the sevoflurane or to a manufacturing process for sevoflurane in which there was sufficient residual water to fall within the claim. Abbott submits that in relying upon the points which I have summarised above, Baxter adopt an over-rigid approach to construction, a proper purposive approach leading, inevitably I think, to the conclusion that all that is required is to make the Lewis acid inhibitor available to inhibit the Lewis acid which will otherwise degrade the sevoflurane.

As I understood the evidence, a liquid Lewis acid inhibitor mixed with the sevoflurane has certain advantages in this regard. First, it can inhibit the action both of a solid Lewis acid such as the activated aluminium oxide of Example 1 of the specification, and it can also inhibit any liquid Lewis acid that might for some reason be present. The class of Lewis acids is very large, and it was not in fact suggested that any particular Lewis acid was likely to turn up accidentally in sevoflurane manufactured and packaged by Baxter. So the question is whether the lining of the container can be viewed as a plausible Lewis acid inhibitor where the Lewis acids concerned are solids. This question gave me much concern, and I asked Professor Kilburn a series of questions from transcript 786. His answers indicated to me that if the Lewis acids in question are solid (and these are the only class of Lewis acids with which the specification is, on the face of it, concerned) a complex mechanism would have to be hypothesised if the lining of the container were to function effectually as a Lewis acid inhibitor. This is only to be expected, because, since both Lewis acid and Lewis acid inhibitor were solids, the opportunity for reaction is more or less non-existent without some complex mechanism taking place between them: he suggested a number of possible mechanisms, but recognised that we were discussing mechanisms which we did not know existed concerning an epoxyphenolic resin of which we were unaware of the properties.

In no sense could the type of mechanism contemplated by Professor Kilburn be viewed as involving compounds that had been combined, without giving to that word an entirely forced meaning. Dr Lessor was challenged on this issue because he had, during the development of the Baxter container, made a number of observations concerning containers with Lewis base characteristics. The first was an invention record dated 10^th January 2000 as follows:

“Concept: PET as a packaging material may inherently provide neutralisation/adsorption of dissolved acids (particularly Lewis acids) which are responsible for the decomposition of anhydrous sevoflurane.

PET is a polyester material which contains RCO₂R linkages. These carboxylate linkages display carbonyl and ester oxygen atoms with available lone pairs of electrons on their surface. Because these oxygen lone are more electron-rich than the oxygen and fluorine atoms present in the sevoflurane molecule, they may preferentially form complexes with any soluble Lewis acid present in liquid sevoflurane which contacts the PET material, thus reducing or eliminating decomposition of the sevoflurane induced by the Lewis acid.

This concept is potentially extensible to any other polymer which displays electron-rich, lone-pair displaying functional groups on its surface, either as part of the backbone of the polymer, as pendent groups on the polymer, or in the form of an additive (either a copolymer additive or non-covalent additive to the polymer matrix).”

The test that is suggested is to contaminate anhydrous sevoflurane with soluble Lewis acids such as aluminium chloride, zinc chloride, hydrogen fluoride and the like, then packaging the contaminated material in PET containers and glass containers as control. The concept here is clearly the use of PET as a (dissolved) Lewis acid inhibitor.

Dr Lessor also proposed that the epoxyphenolic liner would be capable of deactivating Lewis acids that were present in the liner as an embedded impurity - for example, a residual polymerisation catalyst. Dr Lessor says that his primary concern in this email was to assuage the worries of the recipients that the epoxyphenolic might itself be a Lewis acid, and that seems to be quite consistent with the document itself.

While, therefore, it is clear that a liner may act as an inhibitor for dissolved Lewis acid, there is really no evidence at all that a liner would be effective in any way against an insoluble Lewis acid of the kind with which the Patent appears to be concerned.

Against this background I have come to the conclusion that the word “combining” means combining homogeneously and does not include contacting the sevoflurane with solid potential Lewis acid inhibitors. The word “combining” is obviously employed in the claim because of the number of different manners in which water can be introduced into the sevoflurane, as described in [0027] to [0032]. The whole thrust of the specification is that the inhibitor should accompany the Lewis acid after the two are combined, the inhibition of any existing Lewis acid in the sevoflurane of course taking place at the moment of combination – see [0032]. I am also of the view that the use of a protective lining in the container is, as it were, a partial protection only, given that if the sevoflurane is anhydrous and if it subsequently comes into contact with appreciable quantities of said Lewis acid when in use, there is no evidence (speculation apart) that the liner will protect.

Having thus discussed the issues of construction that arise, I can turn to the question of infringement, or more accurately, the question whether the PPD exclusively describes materials which will not infringe.

In consequence of the problems with vaporisers to which I have referred, and consequential concessions made by Abbott, one starts with a water content of not more than 130 ppm, the total impurity limit being not more than 300 ppm. If I am right in my construction of Claim 2, then this material does not infringe, because the water content is too low. If I am wrong, then the question is whether and to what extent Lewis acid inhibitor(s) are present in a quantity sufficient to prevent degradation. If I construe the words “to prevent degradation” in the manner for which Abbott contend – that is, merely to give a better performance than would otherwise be given by anhydrous material – then the description is insufficient, but so is the invention defined by the claim, since the nature of the “insult” appropriate for the test is undefined.

The next matter is the container. I have explained why I do not consider that the lining of a container is combined with its contents. If this is wrong, and the lining is combined with the contents, the protection against solid Lewis acids is, on the evidence, hypothetical. There can in those circumstances be no question of “prevention”, and the resin is not, whether provided in combination with 130 ppm of water or not, sufficient to prevent degradation by solid Lewis acids.

The next matter is paragraph 2 of the “further characterising” part of the PPD. This states that the product in its container “will degrade after 24 hours at 55° C, as indicated by total impurity levels measured by gas chromatography of greater than 300 ppm, pH values less than 4.0 (when challenged with the amounts and type of Lewis acid set out in the Pharmaceutical Development Report … (Annex C))”. No question of relying upon any experimental results arises here. This is a characterising feature of the PPD, and the only question is whether a product which does this has a Lewis acid inhibitor in an amount sufficient to prevent degradation of said quantity of sevoflurane by a Lewis acid. I fail to see how a pH value less than 4.0, which must, in part, indicate the presence of hydrogen fluoride, combined with a total impurity level by gas chromatography of greater than 300 ppm, can be viewed as a material in which degradation has been prevented. Since the accelerated test (24 hours at 55° C) seems to be little different from the range of elevated temperatures employed for comparable tests in the Patent, and since no criticism was advanced of the employment of such conditions in an accelerated test, I conclude that any product having this characteristic cannot be said to infringe. The actual tests employ the same level of Lewis acid as Example 1 of the Patent, and no criticism was directed to the specific total impurity level selected as not being indicative of degradation. I conclude that there is here no infringement, obviously so.

I wish to make a particular observation about the approach adopted by Abbott in this case. I have not yet discussed the fact that Abbott were provided, on 12^th July 2005, with sample-filled containers as proposed to be manufactured and sold by Baxter – these samples being referred to in the PPD itself. Once it is clear that in response to the only challenges specified in the patent in suit in Example 1, properly scaled so as to be appropriate for 250 ml containers, there is degradation on this scale, then the evidential burden shifts. While Abbott are entirely correct that the legal burden of demonstrating non-infringement lies upon Baxter, and while they are equally correct that the decision in Mallory v. Black Sivalls [1977] RPC 321 establishes that, where a description alone is relied on, that description must be completely precise and such as will enable the court to say that no article corresponding to the description could infringe the patent, nevertheless the availability of a sample referred to in the description and studiously ignored by the patentee enables the court to take a fairly robust attitude to the sort of contention that involves suggesting that the claimant’s material, as described, might be immune to the challenges presented by Lewis acids when the material is available for showing conclusively that it does do so. In the present case, there is not even any counterclaim for infringement and no contention is advanced that any material falling within the description does in fact infringe. If this jurisdiction, which is a valuable jurisdiction, is to fulfil its purpose, patentees must, I think, understand that to run a case on what is in effect a case of obstructive non-admission is unhelpful. In an attempt to elicit some sort of case on infringement, I directed that Abbott provide a statement of case in relation to infringement, which failed more or less completely to achieve its object. Much of it was devoted to the issue on amines to which I have referred above which dropped out of the proceedings, and the remainder was given over to a discussion of the function of the epoxyphenolic resin which concluded that the water, the epoxyphenolic resin and the amines, when present together, “have the capacity to inhibit (in whole or in part) the degradation of sevoflurane by Lewis acid”. This document was, in my judgment, as close to being uninformative as it was possible to get. It does, I suppose, point up the issue on construction and on sufficiency of disclosure very concisely by employing the phrase “in whole or in part”. But I view the claimant’s provision of samples to be entirely correct in the circumstances and a step which should have shortened the proceedings. I have considered whether I should not make a declaration of non-infringement in respect of all packaged sevoflurane not substantially different from the samples that were provided to Abbott and ignored, but in the result, that is unnecessary.

Finally, I wish to emphasise that unless the issue of infringement can be decided by reference to a sample of the material, together with a description of how it is made, then the specification is likely to be insufficient. But I make no finding on that basis.

The validity of the patent in suit is challenged on the basis that there has been prior publication of the invention in two published European patent specifications, EP 0701985 (“Kawai”) and EP 0700888 (“Yoshimura”). A great deal of time was taken up with the allegation of prior use by Abbott’s own Japanese supplier, Maruishi Pharmaceutical Co Ltd, including the sale in Japan of 1,200 bottles from a lot, 3430. By a late amendment, I permitted Baxter to raise a very large number of other sales by Abbott to the public between April 1995 and 3^rd January 1997, but in the event Abbott complained that the task of investigating the circumstances of these sales was too great and that their ability to prepare the action would be significantly impaired, were that list of alleged prior sales gone into at the hearing before me. I accordingly ordered that that question be stood over for a subsequent hearing if necessary. I did not revisit my decision to permit the amendment, essentially for two reasons: first, that the public interest involved a proper investigation of prior uses by a patentee; and, second, that the Court of Appeals for the Federal Circuit invalidated the United States equivalent of this patent because the claim construed in the manner that I have rejected covered the sales made by Abbott in the United States before the grace period. Insufficiency is alleged in respect of the invention specified by Claims 2 to 4, and an explicit allegation is made that if Baxter’s method of manufacture infringes Claims 2 to 4, then the specification of the Patent does not disclose the invention claimed clearly enough or completely enough for it to be performed by a person skilled in the art, in that the scope of protection of those claims is so wide that the claims extend beyond the patentee’s contribution to the art – so-called Biogen insufficiency.

Finally, there is a specific point going only to the invention as claimed in Claim 1 of an extension of the disclosure of the specification compared to the application as filed. The matter alleged to have been introduced is the reference at page 3 line 17 of the specification, and in Claim 1, to the absence of soda lime, a material which is not mentioned in the application for the Patent.

Kawai is concerned with inhibiting the accumulation of the impurity fluoromethyl 1,1,3,3,3-pentafluoroisopropenyl ether (compound A) during the distillation of sevoflurane. It is concerned with the distillation of crude sevoflurane during its manufacture in the presence of certain compounds which are referred to as “decomposition suppressive agents”.

The disclosure of Kawai commences, as usual, with an abstract on the front page describing the invention in the following terms:

“A method of purifying fluoromethyl 1,1,1,3,3,3-hexafluorisopropyl ether by distilling the crude ether in the presence of a compound selected from among hydroxides, hydrogenphosphates, phosphates, hydrogencarbonates, borates and sulfites of alkali metals, alkali metal acetates and phthalates, and boric acid. According to this method, a highly pure fluoromethyl 1,1,1,3,3,3-hexafluorisopropyl ether to be used as a medicine, especially as an inhalation anaesthetic, can be produced by effectively inhibiting the decomposition of the ether in the distillation step.”

The distillation described in Example 2 is shown in Table 2 to be entirely free of compound A and to yield very pure sevoflurane, the main distillate having a purity of at least 99.9% and a yield of about 70%. At the end of the description, the following passage occurs under the heading “Industrial Applicability”:

“According to the method of the present invention, it is possible to effectively suppress the decomposition, at the time of distillation, of [sevoflurane] that is used as a pharmaceutical and particularly as an inhalation anaesthetic and thus to obtain [sevoflurane] of high purity.”

It is common ground that after this distillation, the sevoflurane will be saturated with water. The result of the distillation will be an azeotrope in which there is a water phase and a layer of sevoflurane saturated with water, that is, sevoflurane with the maximum possible content of water, 1400 ppm. There is a clear indication that the distillate has a purity better than 99.9%, and in the first example a purity of 99.995%. These figures are obtained by gas chromatography, which will not detect volatile inorganic impurities coming over with the desired distillates.

This raises a question of construction. The claim is to “an anaesthetic composition”. Would the skilled person consider that the sevoflurane saturated with water, with a high degree of purity specified in Table 2 of Kawai, was such a composition? If he would, this is an anticipation. Abbott submit, however, that so far as Claim 1 is concerned, the product of Kawai is not “an anaesthetic composition” because those words require that the material be suitable for administration to human beings. The point does not arise in relation to Claims 2 to 5, because those claims are merely claims to a method of preventing degradation by Lewis acid of a quantity of sevoflurane by the specified steps, through which any manufacturer who employs Kawai’s method of distillation to purify the sevoflurane will necessarily pass. I shall consider the position of Claims 2 to 5 below, and deal with the construction point on Claim 1 first. Normally, words such as “an anaesthetic composition” in this claim would not be considered as limiting the scope of a claim, but rather as describing what is in fact claimed. Thus, the mixture of sevoflurane and Lewis acid inhibitor called for by Claim 1 is an anaesthetic composition. It is plain from the specification as a whole that the invention is contemplated to extend to storage of bulk quantities of sevoflurane (see [0032]) and that it is not concerned only with the final packaging of the material. Indeed, the invention appears to contemplate a composition containing as little as 98% w/w sevoflurane – see [0024]. My preference is for a construction which treats this claim as a claim to the mixture and accordingly covers the product of the distillation in Kawai.

If I am wrong in this, and the words “an anaesthetic composition” do import some sort of requirement that the material be generally suitable for administration to humans, it plainly cannot be limited to material satisfying any particular specification, because, for example, stability would be entirely irrelevant. If the material were then and there suitable for an extemporaneous anaesthesia, that would be good enough.

That point disposes of the concerns expressed by Professor Chambers, which I accept may be well founded, that the skilled person would be concerned as to the high water content of the product of Kawai. The high water content would be a concern because of appearance if the azeotrope separated and because of the potential threat of hydrolysis to produce compound A. On the construction for which Abbott contend, both these factors seem to me to be irrelevant. The material is nonetheless suitable for administration to a human being if it is cloudy, or that it may hydrolyse to produce impurities in the future. Compound A is not present in the material upon its production, as the gas chromatography analysis makes clear, and so from this perspective the product of Kawai is indeed an anaesthetic composition.

The real problem is caused, as I have indicated above, by the potential volatile inorganic impurities, chief among which is hydrogen fluoride. Professor Chambers performed a calculation suggesting that there was a risk that hydrogen fluoride would be carried over. The risk is not said to be great and, in my view, the specification contemplates that the distillation is a final purification step: under the heading “Background Technology” the following passage occurs:

“Hitherto, [sevoflurane] has been widely used as a safe inhalation anaesthetic.

This [sevoflurane] can be obtained, for example, by reacting together fluoromethyl-1,1,1,3,3,3-hexafluoroisopropyl alcohol, formaldehyde, and hydrogen fluoride. The thus produced crude [sevoflurane] contains various by-products. A means is selected to remove these by-products by passing the reaction product through usual treatment steps, that is, steps such as washing with acid, washing with alkali, washing with water, distillation and the like.

However, in the distillation step of this [sevoflurane], it was found that crude [sevoflurane] decomposes or disproportionates and thus can not be distilled and that impurities thus increase. In other words, it was found that defluorohydrogenation of [sevoflurane] occurs during distillation and that [compound A] is gradually formed as a new impurity. This [compound A] which is a decomposition product of [sevoflurane] is a volatile analogous compound of [sevoflurane], and can not be separated because their boiling points are nearly the same. It is needless to say that the contamination of the product thereby is extremely unfavourable in use as an inhalation anaesthetic, and thus an immediate solution has been desired.”

This seems to me to amount to clear and unambiguous instructions to use “the usual treatment steps” to remove existing by-products, of which hydrogen fluoride will be one. This is not a question of obviousness; these are clear directions.

I am conscious that this point was not argued, and in any event Professor Chambers accepted that if one was concerned about the presence of hydrogen fluoride, one would wash with caustic, for example, before the ultimate drying step with which he would proceed because of his concerns about stability in the light of a possible hydrolysis. So there are either clear and unambiguous directions to remove acid by-products by caustic wash before the distillation or it is overwhelmingly obvious to wash the material before drying it. In point of fact, it is common ground that the material purified of hydrogen fluoride and saturated with water is in fact an anaesthetic composition, and it is inevitable, in the correct sense of that word, that any preparation of pharmaceutically acceptable sevoflurane which employs the distillation method disclosed by Kawai will pass through a stage at which it is a composition within the claim. So Claim 1 is anticipated. So too are Claims 2 to 5.

I do not think this question really arises. I am, however, astonished that there is a suggestion that the skilled person would employ the method of Kawai and yet end up with a product (prior to drying) that contained any pharmaceutically unacceptable contaminant. If it is obvious, as Professor Chambers acknowledged that it was, that the drying step – if employed – was the last step, then the point on obviousness really becomes marginal. However, it was contended also on behalf of Baxter that a drying step, which adds a further process step, would not be employed in production unless stability studies indicated that the sevoflurane produced, containing 1400 ppm water or 0.14% w/w, was in fact unacceptable from the point of view of stability. Dr Woodhouse, who gave evidence on behalf of Baxter, and had wide experience of the formulation of inhaled drugs during his time with Glaxo, took the view that the presence of water would certainly be investigated, having regard to the fact that there was on sale an established sevoflurane product from Abbott which contained much smaller amounts of water. His view was that no pharmaceutical chemist would remove the water unless it was a problem. He accepted that the reference in the literature to hydrolysis of sevoflurane was correct, but he took the view that the statement in the same paper that sevoflurane was stable without additives for over one year at 45° C in amber glass bottles indicated that there was less of a problem than might appear. He said that what one would do is to conduct normal stability tests, about which he gave detailed and unchallenged evidence. I confess that I found Dr Woodhouse’s description of product development of a commercial sevoflurane formulation in the light of Kawai to be quite compelling. He made it clear that whether or not there was another product (he called it the originator product) on the market, every producer would have to perform its own stability studies, and so there was a good reason for investigating the stability of the Kawai product, further washed, if need be, to remove hydrogen fluoride. He fairly accepted that the Wallin paper upon which Professor Chambers based his concerns provided a warning light, but considered that the inevitable stability testing would deal with it.

The Windsurfing analysis inevitably requires me to consider the distinction between the disclosure of Kawai on the one hand and the inventive concept of the patent in suit on the other. So far as Claim 1 of the patent in suit is concerned, the distinction, if any, lies in Kawai’s failure to be completely explicit about processing subsequent to the distillation step. It does not say what processing steps are to take place, if any, after the distillation. Since I accept Professor Chambers’ evidence that the drying step will be the last step in production, and since I accept that at some stage it is obvious that volatile inorganic impurities, particularly hydrogen fluoride, will have to be removed, conventionally by washing, the only question is whether Kawai implicitly discloses a material which will be subsequently dried. For the reasons given by Dr Woodhouse, I consider that a drying step will be omitted if stability studies demonstrate that it is not necessary. The material is in fact known to be stable when saturated with water, and it is accordingly obvious to make a compound within Claim 1. The result is that Claims 2 to 5 are equally obvious if I am wrong in my view that Kawai in any event discloses a material as being suitable for use as an anaesthetic.

The allegation here is twofold. Central Glass’s lot 211231, which consisted of 1,611 kg of sevoflurane, was supplied by Central Glass to Maruishi. This quantity amounts to about 4,200 bottles, and was divided into two lots by Maruishi: lot 3426 and lot 3430. What is said is that part of each of these lots was shipped to Abbott to be used in clinical trials in the United States, and the remainder were, it is contended, disposed of in the ordinary course in Japan. The argument proceeds in stages, and I shall attempt to summarise its essential features here. Annex 1 of the Grounds of Invalidity is a certificate of analysis of a lot of sevoflurane, lot 3430, produced and disposed of by Maruishi in 1993. The water content is stated to be not more than 0.2% w/v – that is weight for volume and is about 1326 ppm – and the actual analysis is 0.07% w/v or about 464 ppm. It will be recalled that 464 ppm is within the preferred range of water contents identified in [0027] of the Patent. The certificate of analysis is dated 20^th May 1993 and states that lot 3430 was manufactured on 30^th April 1993. The other lot having a high water content was identified on the documents: lot 3426, which is said, like lot 3430, to consist of 1,200 bottles of 250 ml each, was stated on the certificate of analysis dated 20^th May 1993 to have a water content of 0.07% w/v as well.

As I have said, both lot 3426 and lot 3430 were derived from the same bulk lot supplied to Maruishi by the manufacturers, Central Glass. For the purposes of their application for a new drug to the US Federal Drug Administration (“FDA”) Abbott were obliged to set out the bulk drug and final product lots of sevoflurane used in the development, and it is confirmed that the bulk lot from which both 3426 and 3430 were obtained was bulk lot number 211231. All the final dosage forms were manufactured by Maruishi.

The evidence here was given on Baxter’s part by Dr Lessor, who worked off the documents. Ms Delgado-Herrera gave evidence for Abbott. She was a poor witness, tending not to answer difficult questions and also, I think, far too conscious of where Abbott’s interests lay. I refer here in summary only to some concessions she was prepared to make. I found Dr Lessor’s deductions to be generally convincing.

It appears that the total size of bulk lot 211231 was 1,611 kg. Dr Lessor calculates that this represents approximately 4,000-4,250 ml bottles and suggests that both lot 3426 and lot 3430 would have been substantially bigger than the 1,200 bottles to which the certificates of analysis relate. Indeed, he suggests, basing himself on Table XXXVI of Abbott’s New Drug Application filed with the FDA, that batch 3426 in fact consisted of 1,914 bottles in total. The contention is that the balance of bottles in these batches would have been retained by Maruishi and sold to other customers in Japan or elsewhere where such sales were permitted. With a view to obtaining information from Maruishi, Lewison J granted letters of request, but no response had been received by the time the trial before me opened. The evidence from Ms Delgado-Herrera was that the relationship between Abbott and Maruishi was good, and so the position on the evidence is that Baxter had done what they could to obtain the evidence and Abbott, unsurprisingly, did nothing. In these circumstances the question is whether there is any evidence to throw doubt upon the inference which Baxter seek to draw from these bald facts, that the balance of the two lots was indeed retained by Maruishi. Ms Delgado-Herrera says, firstly, that the analysis issued by Central Glass in relation to bulk lot 211231 was less than 30 ppm, and that that represented the water content immediately prior to shipment to Maruishi. Secondly, she appears to have accepted in her first witness statement that Abbott were supplied with 1,200 bottles from each lot. She maintains, however, that the whole of the two Maruishi lots 3426 and 3430 were sent to Abbott rather than the 1,200 bottles from each lot specified in the analysis certificate. This would of course depend upon there having been an order for more than 1,200 bottles in each lot, but there is simply no plausible indication that this happened. Furthermore, the FDA contains a very complete record of all the experimentation that was done on this material, and if the balance of the two lots had been supplied to Abbott in the United States, there is neither an analysis certificate nor any trace of any work done using them, and the NDA is silent. Indeed, there are powerful reasons for supposing that only 1,200 bottles from each batch were received. Ms Delgado-Herrera had the opportunity to check all the clinical trial documents and was able to account for nearly all the 2,400 bottles that Baxter say were supplied from the two lots, but no more than this. The second relevant consideration is that neither of the two lots was as much as 2,400 in size, and so it makes good sense to supply half the notional order with sevoflurane from two lots, both of which derive from the original bulk lot.

There was no evidence that Abbott invariably received whole lots from Maruishi. While, therefore, there is a clear paper trail in relation to the 2,400 bottles, there is simply no documentary indication at all as to any other bottles forming parts of lots 3426 and 3430, had Maruishi supplied them. I do not feel that it is an extravagant inference to say that this evidence points, on the balance of probabilities, to Abbott having been supplied with 1,200 bottles from each of the two lots originating from a common bulk lot by Maruishi, and that Maruishi retained the remaining bottles in each lot, however many of those there may have been. It inevitably follows that the inference must be drawn that Maruishi disposed of those bottles in the ordinary course of business, and that inference can only be rebutted by evidence from Maruishi. As I have indicated, Maruishi were reluctant to respond to the letters of request, and Abbott have sought no assistance from them. I conclude that these bottles were sold on the Japanese market. I conclude further that this sevoflurane contained water at a concentration of 460 ppm or thereabouts.

There is one matter to which I should refer. As I have already indicated, Table XXXVI of the new drug application shows that lot 3426 was a lot containing 1,914 bottles. A lengthy cross-examination of Ms Delgado-Herrera, together with her calculation as modified and exhibited at X6, satisfied me that the entry in the table cannot relate to the number of bottles used and must, in fact, relate to the number of bottles in the actual lot. So Table XXXVI does not suggest, when properly understood, that Abbott used 1,914 bottles of lot 3426 for clinical trials.

The sales to Abbott from lots 3426 and 3430 were of sevoflurane undoubtedly used for the purpose of clinical trials. Were Abbott free to disclose the water content of this material without regard to its suppliers, Maruishi? The approach of the parties is different to this extent. Mr Floyd QC, on behalf of Baxter, submits that the material was on open sale in Japan, the water content of the material made available to Abbott in the United States made the water content available to the public because Abbott were free to analyse the material and announce the results. Mr Thorley QC, for Abbott, submits that Abbott, Central Glass and Maruishi were cooperating to obtain regulatory approval of sevoflurane worldwide and that, properly viewed, the studies in the United States are part of this endeavour. An agreement between Maruishi and Abbott provides that information exchanged between them would be treated as confidential, and this obligation relates to the water content of the material being tested in the US, whether or not such material is also being sold in Japan. After all, without disclosure of the supply arrangements, there is no reason why the public should know that the material being used in the US is the same as that in Japan. This contention does, contrary to Mr Floyd’s submissions, obtain some support from a passage in the cross-examination of Dr Lessor, where he accepted that, had he been a scientist or clinician involved in Abbott’s trials and measured the moisture content of the sevoflurane at 400 ppm, noticing that it was higher than that which was on sale in Japan, he would not have considered himself free to publish that information without the consent of Abbott or Maruishi. Mr Thorley also observes, with some justification, that Baxter has designated the water levels in its commercially available sevoflurane as confidential in these proceedings. Now, it is clear that Dr Lessor’s answer to Mr Thorley was predicated upon his having noticed that the water content of the material being tested was higher than that available in Japan – always assuming that anybody saw any significance in the water levels in any event. On the whole, I consider that the question is really decided by the terms of the agreement between Maruishi and Abbott, and I conclude that Abbott were not free to disclose this information.

For the foregoing reasons, I consider that the evidence indicates on balance that bottles from lots 3426 and 2430 were retained by Maruishi in Japan, and the inevitable inference is that they were sold in Japan, thereby making sevoflurane having a water content of 400 ppm available to the public in Japan, but that there is no relevant non-confidential use in the US of the material used in clinical trials. The result is that Claims 1 to 5 are anticipated.

A separate head of prior use is advanced on the basis of an information disclosure statement lodged by Abbott in the course of its prosecution of the patent corresponding to this patent in the United States. US law requires a good-faith disclosure by the applicant of material relevant to validity known to it at the time of the application. Failure to make a good-faith disclosure may result in an unenforceable patent. The Information Disclosure Statement (“IDS”) is Annex 3 to the Grounds of Invalidity in the present case and it specifies a table headed “Sevoflurane Water Content” in which a number of batches have water contents comfortably in excess of the level identified by the Patent as corresponding to an anhydrous sevoflurane, and in one case containing 131 ppm of water.

As I have construed the claim, none of these materials are capable of anticipating the claim. I will not, therefore, discuss in detail whether, and to what extent, actual water contents are made available. If the material is sold on the open market, the claim is good. The question is difficult, since it involves asking questions about the claim which cannot sensibly be answered, for reasons I discuss below. I cannot, however, accept that this is such a case as that discussed in G02/88 Mobil, itself a notoriously difficult case to analyse. The present case is to be distinguished from Mobil purely on the ground that no material within the present claim necessarily protects against insult – it all depends on the insult. The only novelty in this claim is new information about a known use, and that cannot save the claim from anticipation. If I am incorrect in my construction, the claim is invalid for reasons that I have already referred to and which I shall discuss under the heading “Insufficiency” below, and I consider that the question whether or not the quantity of water is sufficient to “prevent degradation” is incapable of being answered, since the magnitude of the “insult” is unspecified either in the Patent itself or as a matter of common general knowledge. There is no question of there being a “predictable” amount of Lewis acid which might be present. There was no evidence of any predictable amount of Lewis acid, contrary to Mr Thorley’s submission. It is therefore unnecessary for me to consider the position, were not the claim on its proper construction to require a minimum water level of 150 ppm.

As pleaded, the objection of insufficiency, which is restricted to Claims 2 to 4, is as follows:

“The Patent provides no or no sufficient disclosure of what amount of Lewis acid inhibitor is sufficient in any given case to prevent degradation of any given quantity of sevoflurane. In particular the Patent does not address the amount of such inhibitor to be added for different types of Lewis acid, for different types of container for the product, for different amounts of the product, for different amounts of Lewis acid, for different methods of producing and storing the product or otherwise or how such amounts of such Lewis acid may be determined without undue research, inquiry or experimentation.”

I have really covered all this material in attempting to put a reasonable interpretation on the claims in their context in the specification. I have rejected the contention that the claim is to be given a comparatively straightforward meaning (“enough water that there is less degradation than there would be without”) because it is inconsistent with the use of the word “prevent” and because it leaves the skilled man with no guidance at all as to the magnitude and nature of the “insult” to which his sevoflurane is to be subjected if it is not the equivalent of 10 mg, 20 mg and 50 mg activated alumina in 20 mls sevoflurane. No guidance is given as to how the skilled man might effectively estimate the possible future insults to be undergone by his sevoflurane, particularly when such insults are unpredictable and exceptionally infrequent. If, therefore, the claim is to be construed as requiring a functionally defined quantity of water – an “effective amount” – then I find it impossible to work out how to define a suitable test. If sevoflurane were inherently unstable, the position might well be otherwise, but that is not the case. Professor Kilburn suggested a statistical analysis of packaged sevoflurane. But this takes one nowhere, because the material absent an unusual event is stable. Thereafter, the matter becomes speculative. I think possibly one proper view that can be taken of the problem is that in the event that an event occurs resulting in Lewis acid contamination of production sevoflurane, a possible approach is to include within the sevoflurane a prophylactic quantity of water suitable for preventing degradation caused by such an event. But the Patent is unhelpful on suitable events, unless one takes the equivalents of Example 1. Professor Kilburn identified rust (ferric oxide, Fe₂O₃)as a potential Lewis acid appropriate for use in this context. Rust is not mentioned in the Patent, but it was in fact the Lewis acid which causedthe problems leading to Abbott’s employment of water as a Lewis acid inhibitor in their sevoflurane. Professor Kilburn said this:

“Well, as I said yesterday, there are other experiments that you can do. You can actually set about establishing what is the minimum amount of Lewis acid you have to worry about. That is not something which is described in that paragraph. But, as I said before, there will be – let us take rust as an example. You can subject your sevoflurane to varying amounts of rust with an appropriate control group as you see fit. You will establish that for some amount of Lewis acid, there will be a level of degradation, but not a level that you will have to concern yourself with, which, for me, would be, as according to the Patent that you have to be sure, below 669 ppm degradants. So you will establish that, for example, maybe 0.1 mg of rust you do not need to worry about, but, say, 5 mg you do need to worry yourself with. You can then do experiments with the 5 mg with varying amounts of water to establish what is the appropriate level of water you need to guard against that level of degradation.”

I have already explained why I do not accept that the Patent considers a level of impurities below 669 ppm to be acceptable. For the reasons set out above, the Patent makes no such indication, rather the opposite. But I accept Mr Floyd’s submission that such a course of work is, on the face of it, an indication of serious and incurable insufficiency.

I believe that a construction of the claim which I have decided is appropriate avoids the most obvious problems of insufficiency. If I am wrong, Claims 2 to 4 are hopelessly insufficient.

Finally, I must deal with the objection of added matter, which is both short and straightforward. Claim 1 is limited to a composition which contains no soda lime. The description of the invention in [0008] is as follows:

“The present invention involves a stable anaesthetic composition that contains a fluoroether compound having an alpha fluoroether moiety having added thereto an effective stabilizing amount of a Lewis acid inhibitor. The fluoroether compound is sevoflurane and the Lewis acid inhibitor is water. No soda lime is present in the composition. The composition can be prepared by adding the Lewis acid inhibitor to the fluoroether compound, by adding the fluoroether compound to the Lewis acid inhibitor, or by washing a container with the Lewis acid inhibitor and then adding the fluoroether compound.”

Section 72(1)(d) Patents Act 1977 provides that a patent may be revoked upon the ground that “the matter disclosed in the specification of the patent extends beyond that disclosed in the application for the patent, as filed, or, if the patent was granted on a new application filed under Section 8(3), 12 or 37(4) above or as mentioned in Section 15(4) above, in the earlier application, as filed”.

It is common ground that there is no reference to soda lime in the application. So the statement that “no soda lime is present in the composition” is new matter. While I can envisage that it might be possible to amend to remove this statement in the body of the specification, it cannot be removed from Claim 1, where it appears, and its deletion from which would extend the protection conferred by that claim contrary to the provisions of s.76 of the 1977 Act. The same limitation does not appear in the method claims, and [0009], which describes the invention so far as it relates to method, does not contain this qualification.

I am at present of the view that the objection is made out, and that the only possible curative amendment is deletion of the offending words in [0008], together with deletion of Claim 1. At the moment, I cannot see why matter added to the disclosure which relates only to one aspect of the invention should, after deletion, continue to affect the validity of claims that make no reference to the added matter, and are adequately supported (subject to my observations above) by the remainder of the disclosure. Without such an amendment, however, it seems to me clear that the Patent must be revoked upon this ground also.

In the light of my conclusions above, I do not consider that it is productive to consider so-called “Biogen” insufficiency in the context of this patent. I have already mentioned that certain allegations of prior use remain to be tried, but whether that is necessary must be decided hereafter.

The Patent is invalid and must be revoked in its entirety. Had the Patent been valid, I would have granted the declaration of non-infringement sought.