Gedeon Richter Plc v Bayer Pharma AG

This is an appeal against a decision of Floyd J in a patent action concerning inventions in the field of oral contraceptives.

The appellant, Gedeon Richter plc (“Richter”), is a generic pharmaceutical company. At trial it sought revocation of two patents owned by the respondent, Bayer Schering Pharma AG, which was formerly known as Bayer Schering Pharma AG (“Schering”): EP (UK) Nos. 1,380,301 (“301”) and 1,598,069 (“069”). They were granted on divisional applications based upon the same parent application WO 01/15701 (“the parent application”).

Both patents relate to immediate release formulations of the steroidal hormones drospirenone (“DSP”) and ethinylestradiol (“EE”). DSP is a gestagen and EE is an estrogen. In such a formulation contraceptive reliability is mainly provided by the gestagen and the estrogen acts to increase the ovulating inhibitory effect of the gestagen and to ensure cycle stability.

The judge held:

On this appeal, Richter challenges findings (b), (c) and (d). The judge refused permission to appeal against his findings of non-obviousness. Lord Neuberger of Abbotsbury MR subsequently granted permission to appeal against these findings, albeit with “considerable scepticism” as to the likely success of any such appeal.

Formulation of an oral contraceptive is critically important. The dosage must be such that the contraceptive is wholly effective at inhibiting ovulation. On the other hand, too high a dosage may result in undesirable side-effects.

In this regard, DSP presented particular challenges. First, it is poorly soluble, which carries with it the risk of poor absorption. The skilled team would have known of various techniques for dealing with this problem, two of which, micronisation to reduce the drug particle size and application of the drug to very small inert carrier particles, are expressly discussed in the parent application. The second is that DSP has a tendency to change into an inactive isomer under the acid conditions found in the stomach. It was also well known that steps taken to increase the solubility of such a drug by, for example, micronisation, would be likely to increase the rate of degradation.

The judge described the common general knowledge in relation to these issues at [25] – [26]:

The parties were agreed that during the pre-formulation testing stage a pharmaceutical formulator would subject a new candidate active pharmaceutical ingredient (“API”) to a series of in vitro tests in order to establish its solubility and stability in different environments. Should these tests reveal the API to be acid labile, the formulator would know that it is possible to protect it by using an enteric coat, as the judge explained at [31] – [36]:

With this background we can now turn to the particular issues arising on this appeal.

The test for added matter was explained by Aldous J in Bonzel v Intervention (No 3) [1991] RPC 553 at 574; by this court in Vector v Glatt Air Techniques [2007] EWCA Civ 805, [2008] RPC 10 at [4] – [9]; and again by this court in Napp Pharmaceutical Holdings v Ratiopharm [2009] EWCA Civ 252, [2009] RPC 18 at [71] and [98] – [99].

Mr Simon Thorley QC, who appeared on the appeal on behalf of Richter, particularly emphasised the following aspects of the guidance given in those cases.

First, the task for the court in a case such as this is (1) to ascertain through the eyes of the skilled addressee what is disclosed, both explicitly and implicitly in the application; (2) to do the same in respect of the patent; (3) to compare the two disclosures and decide whether any subject matter relevant to the invention has been added whether by deletion or addition. The comparison is strict in the sense that the subject matter will be added unless such matter is clearly and unambiguously disclosed in the application.

Second, it is appropriate to consider what has been disclosed both expressly and implicitly. The addition of a reference to that which the skilled person would take for granted does not matter. On the other hand, it is to be emphasised this is not an obviousness test. A patentee is not permitted to add matter which would have been obvious to the skilled person from the application.

Third, the idea underlying the prohibition is that an applicant should not be allowed to improve his position by adding subject matter not disclosed in the application, which would give him an unwarranted advantage and could be damaging to the legal security of third parties relying on the original disclosure.

Fourth, it is important to avoid hindsight. Care must be taken to consider the disclosure of the application through the eyes of the skilled person who has not seen the amended specification and consequently does not know what he is looking for. This is particularly important where the subject matter is said to be implicitly disclosed in the application.

Finally, it is important to distinguish between the disclosure of subject matter and the scope of claim. Clearly not everything which falls within the scope of the claim is necessarily disclosed, as this court explained in the Napp Pharmaceutical case at [98] – [99]:

Mr Thorley submitted that the judge fell into error in four ways. First, he failed to emphasise the Bonzel requirement of a strict comparison; second he failed to remind himself of the dangers of hindsight; third, although he reminded himself to draw a line between implicit disclosure and that which is obvious, he did not address himself as to where the dividing line between the two is; fourth, he came to deal with the law on added matter having already considered the disclosure in the parent application and that of the granted patents.

We are unable to accept any of these submissions. Floyd J addressed the law at [71] of his judgment. He set out the relevant test and then referred to the decision of Arnold J in Abbott Laboratories v Medinol [2010] EWHC 2865 (Pat) in which he cited (at [251]-[253]) all the key passages in the cases to which we have referred. Floyd J then emphasised four particular points: the fact that a claim covers something does not mean it discloses it; express disclosure in the patent of that which is implicit in the application does not amount to added matter; nevertheless, implicit disclosure is to be distinguished from matter which would be obvious to the skilled reader; and an eye needs to be kept on “impermissible intermediate generalisation”, as explained by Pumfrey J in Palmaz’s European Patents [1999] RPC 47 at 71.

In these circumstances we regard it as inconceivable that the judge did not have well in mind that subject matter will be added unless it is clearly and unambiguously disclosed in the application explicitly or implicitly, the so called strict comparison requirement. Further, the judge expressly reminded himself of the need to draw a line between implicit disclosure and that which is obvious and so would have been well aware of the dangers of hindsight. As to where the dividing line between the two is, that is a matter which can only be decided on a case by case basis and is something to which we shall return. Finally, we see no reason at all why a judge should be required to summarise the relevant legal principles before considering the disclosure. What matters is that the judge directed himself properly as to the law prior to considering whether there was added matter, and in this case he did so.

The parent application begins with the description of the “Field of the Invention” and explains that it relates to a pharmaceutical composition comprising DSP and EE.

The “Background of the Invention” section of the parent application then explains that oral contraceptives containing a combination of a gestagen and an estrogen have been used since the 1960s and that the gestagen-like activity of DSP has been described in the prior art, as has its use with EE as an oral contraceptive.

That brings us to the “Summary of the Invention” which explains that it has surprisingly been found that a hitherto undisclosed minimum dosage level of DSP is required for reliable contraceptive activity, as has a preferred maximum dosage at which unpleasant side effects, in particular excessive diuresis, may be avoided. Accordingly, the invention is said to provide in one of its various aspects a pharmaceutical composition comprising about 2-4mg of DSP and 0.01-0.5mg of EE, together with one or more pharmaceutically acceptable carriers or excipients. This, it will be seen, corresponds to claim 1 of the parent application.

There then follows a “Detailed Disclosure of the Invention”. It begins (at page 4, lines 4-9) with a description of the lack of solubility of DSP and its tendency to degrade under acid or alkaline conditions. It then explains that the compound is advantageously provided in a form that promotes rapid dissolution to promote good bioavailability:

This is followed (at page 4, lines 11-24) by a description of two particular ways of achieving rapid dissolution, namely micronisation and spraying a solution of DSP onto the surface of inert carrier particles:

This passage also teaches that rapid dissolution is defined by a particular level of performance using the USP dissolution test apparatus (“the USP test”).

Then (at page 4, lines 26-31) the parent application explains that the in vitro dissolution rate of DSP is connected to the dissolution rate in vivo resulting in rapid absorption, a reduction in degradation and, consequently, high bioavailability:

On page 5, lines 11-16, it is explained that the inclusion of carriers or excipients may promote dissolution of both APIs:

Then (at page 9 lines 10-16) there is an important passage explaining how the composition of the invention may be formulated:

We can now turn to the claims and, for the purposes of this appeal, need only recite claims 1, 2 and 7-9:

The judge identified the material passages of the parent application at [38] – [49]. Having referred to the “Summary of the Invention”, the judge said this at [40]:

A little later, having referred to the important passages on page 4, lines 4-31, the judge continued at [46]:

Mr Thorley submitted the judge fell into error at this point in failing to make express findings as to what was explicit and was unambiguously to be implied, particularly in relation to the important passage on page 9, lines 10-16, recited at [29] above.

We do not believe that the judge can fairly be criticised for failing to adopt such a formulaic approach. He set out the relevant parts of the parent application, then turned to the patents, and, having addressed the law on added matter, considered each of the specific allegations. That, it seems to us, was a perfectly permissible approach for him to have adopted. Nevertheless, in light of the criticism advanced by Mr Thorley we shall take the course he submitted the judge should have followed.

Mr Thorley contended that it is plain from the “Summary of the Invention” section of the parent application that the invention has at its heart what is said to be the surprising discovery that a previously undisclosed minimum dosage of DSP is required for reliable contraceptive activity, and a preferred maximum dosage has been identified at which undesirable side effects may be avoided. Against this background he argued that the passage on page 4, lines 4-9, recited at [24] above, teaches that a rapid dissolution form is optional, albeit preferred, but contains no teaching as to how this is to be achieved. There follows, at lines 11-24, recited at [25] above, a description of the “surprising” discovery that rapid dissolution occurs when DSP is provided in specific micronised form and that, as an alternative, it is possible to dissolve the DSP in a suitable solvent and spray it onto the surface of inert carrier particles. There is, Mr Thorley continued, no disclosure that it is possible to obtain such rapid dissolution by any other means.

Mr Thorley then turned to the important passage on page 9, lines 10-16, and argued that the composition of the invention here referred to is the composition that provides 2-4 mg per day of DSP. In context, this passage therefore teaches that such a composition may be formulated in any manner known to the art. But it does not disclose, whether explicitly or implicitly, that rapid dissolution may be achieved by any method other than micronisation or spraying.

We accept that the skilled person reading the parent application would understand it to be teaching that the inventors have made the surprising discovery that a previously undisclosed dosage range of DSP is required for reliable contraceptive activity and desirable in order to avoid unpleasant side effects. However, we have no doubt the skilled person would also understand the passage at page 4, lines 4-9 to be teaching that DSP is advantageously provided in a form that promotes rapid dissolution to ensure good bioavailability. This teaching would excite the attention of the skilled person because of his concern that rapid dissolution, should it take place in the acid environment of the stomach, would cause degradation. He would therefore be interested too in the tentative explanation provided at lines page 4, lines 26-31 that the in vitro dissolution rate of DSP is connected to the dissolution rate in vivo and the consequent rapid absorption of DSP following oral administration.

It is true that the only methods of producing a rapid dissolution form of DSP specifically disclosed in the parent application are micronisation and the spraying of inert carrier particles. But that does not mean to say that the skilled person would understand the parent application to be teaching that these are the only ways that rapid dissolution may be achieved. To the contrary, we believe that the skilled person would understand these particular methods are described by way of example together with a method by which rapid dissolution may be tested, that is to say the USP test. It was common ground he would know there were other ways by which rapid dissolution could be achieved, and we have no doubt that he would understand that he could use these too.

We are confirmed in this view by the claims of the parent application which form part of its disclosure. Claim 1 is directed to the particular dosage ranges of the two APIs, DSP and EE. Claim 7 is directed to a composition which contains a carrier or excipient selected so as to promote rapid dissolution of these agents. Then, importantly, claims 8 and 9 are directed to compositions which achieve rapid dissolution when tested using the USP test. They contain no limitation as to how the compositions are to be made.

The judge summarised his conclusions as to the disclosure of the parent application at [78] in terms which cannot be faulted:

The body of the 301 specification is largely the same as that of the parent application. In the “Background of the Invention” section, paragraphs [0009] and [0010] identify two additional pieces of prior art and the description explains that neither discloses compositions having a rapid dissolution of DSP.

That brings us to the “Detailed Disclosure of the Invention” which, at paragraphs [0016] – [0020], corresponds to the description in the parent application from pages 4-5. Later in this part of the description, paragraph [0038] corresponds to the paragraph of the parent application beginning at page 9, line 10.

As for the claims, we need only refer to claim 1 which, as amended following the judgment of Floyd J, reads:

The invention therefore consists of a tablet comprising DSP and EE in a particular dosage range formulated so as to promote their rapid dissolution as measured by the USP test.

The body of the specification of the 069 Patent is essentially the same as that of the 301 Patent. As we have mentioned, the judge found claim 1 invalid but permitted Schering to delete claims 1-5. Thus this appeal proceeds on the basis of old claim 6, now claim 1. It reads:

Three objections were originally taken against the 301 patent. Two have been met by amendment. One remains. It turns upon the disclosure to the skilled person of the sentence appearing at page 9, lines 10-11 of the parent application and paragraph [0038] of 301:

Mr Thorley contended that the words are the same but the disclosure is different because the invention is different. His argument ran as follows.

In the case of the parent application, the composition of the invention is the combination of DSP in an amount of 2-4 mg and EE in an amount from 0.01-0.05 mg. Thus the sentence at page 9, lines 10-11 is teaching that such a composition may be formulated in any manner known in the pharmaceutical art.

By contrast, in the case of 301 patent, the invention comprises not just the combination of DSP and EE in the defined amounts but also their formulation into a rapidly dissolving composition as defined by the USP test. The same sentence therefore has a different meaning. The skilled addressee is now being told that any method known to him as part of his common general knowledge can be employed to achieve the rapid dissolution advantage of the invention.

Thus, Mr Thorley continued, it is plain that there is added matter. Whereas in the parent application, the only disclosure as to how to obtain rapid dissolution is by virtue of micronisation or spraying, in the 301 patent, the express teaching is that any known method of obtaining rapid dissolution can be used.

We accept that the skilled person would understand from paragraph [0038] that he may formulate the composition of the invention in such a way as to promote its rapid dissolution in any manner known in the art. That is not to say that every method of formulation will necessarily produce a composition meeting the rapid dissolution requirement of the claim. But the paragraph does teach the skilled person that he may use any method known in the art to try and achieve that aim.

In our judgment this does not, however, constitute added matter. It is precisely what is taught and disclosed on pages 4 and 5 of the parent application in the passages to which we have referred. Here it is explained that the dissolution rate of DSP in vitro appears to be connected to the dissolution rate in vivo resulting in its rapid absorption upon oral administration. It is therefore desirable to formulate the composition in a way which promotes its rapid dissolution. Two particular ways of doing so are described, namely micronisation and the coating of inert carrier particles. But these are only provided by way of example and there is no suggestion that they are the only ways in which rapid dissolution may be achieved. Indeed it was accepted that the skilled person would know of others.

It follows the judge was right to say at [76]:

In short – the parent taught generally that rapid dissolution ensured “good bioavailability”; it disclosed two methods of achieving that (micronisation and coating of an inert carrier), and the skilled reader would know of other methods which could be used to try and achieve rapid dissolution. The patent as granted adds no new teaching to that information.

Once again, three allegations were originally advanced but only one remains. And it is very similar to that maintained against the 301 patent.

Mr Thorley contended the invention referred to in paragraph [0038] relates to a composition having DSP on inert particles regardless of how it got there. So, the argument went, the DSP can be put on the surface of the inert particles by any known means. But the parent application does not teach this. There is no disclosure of alternative methods at all, far less the clear and unambiguous disclosure required by the law on added matter.

Just as the allegation is very similar to that advanced against the 301 patent, so too is the answer to it. For the reasons we have given we are entirely satisfied that the skilled person reading the parent application would understand that what matters is rapid dissolution; that one way of achieving rapid dissolution is to use inert carrier particles with DSP on their surface and that one way of getting the DSP on to the surface of the inert carrier particles is to spray them. But it would also be clear to the skilled person that he could use any other known technique for coating the particles. In our judgment the judge was therefore right to hold as he did at [80]:

For all of these reasons we are satisfied that the appeal on added matter should be dismissed.

Below the patents as amended (which it was and is common ground can be considered together for this purpose) were said to be obvious over prior published papers called “Lachnit”, “Oelkers”, “Krause I” and “Krause III.”

The judge rejected all the attacks. Before us Mr Thorley only seeks to resurrect one of these. He contends that the judge wrongly rejected the attack based on a combination of Oelkers and the Krause papers (there is no material difference between them).

Before proceeding further, we draw attention to some dates, for it is trite law that these matter when considering obviousness – the older (from the priority date of a patent under attack) a piece of prior art said to render a patent obvious, the harder it is to show obviousness. The priority date of the two patents is August 1999. Oelkers was published in June 1995 and the two Krause papers in 1982 and 1983. Mr Thorley’s case involves showing that the skilled team in 1999 would read a 1995 document with documents published 12 or 13 years earlier than that. He did not shrink from that.

The judge summarised the disclosures of the documents in a manner not challenged by either side so we gratefully borrow:

Mr Thorley’s case on obviousness runs in a series of steps:

The skilled team (which would, it is common ground, include a skilled formulator and a skilled medicinal chemist) would read Oelkers. It discloses a combined oral contraceptive which would be of interest to the team because it may be of special benefit to women susceptible to weight gain and a rise in blood pressure. Oelckers had actually formulated pills of different doses and tried them with three groups of twenty women against a reference sample.

The team would decide that Oelkers was worth following up. Since Oelkers says nothing about formulation, the follow up would consist of the standard step of in vitro tests to see how soluble and stable the compounds were. The tests would be in what one can call “stomach conditions” i.e. at 37⁰C and stomach acidity. The experts were agreed that they would do the tests themselves rather than see what the literature said about the matter.

The experiments would show that within 45 minutes (within the mean residence time of a drug in the stomach) half the DSP would have isomerised (and thus rendered inactive).

The team would then consider what to do about this. Only two routes were open: one would be to try an enteric coating (the usual method for protecting a drug from destruction by stomach acid so that it could reach the small intestine where it would be absorbed) or to disregard the apparent problem of an uncoated pill at this stage.

The skilled team would also have done a literature search on DSP. This would produce only 11 papers, including those by Krause, concerned with spirorenone (SP).

The skilled team would read Krause and learn that the observed in vitro isomerisation of SP was not replicated in vivo.

So the skilled team would say to itself: “We wonder whether what happens with SP in vivo (admittedly with a macro-crystalline form of the drug and other differences) would also happen with DSP which is structurally very similar to SP. If it also survived in the stomach notwithstanding fast dissolution, then enteric coating (which might have its own problems) would not be necessary. It might work, so let’s try.”

And that would lead them to the inventions of both patents.

The judge did not accept this multi-stage approach. He was prepared to accept stages (1) and (2). Stage (1) (reading a prior art citation) is required as a matter of law. Stage (2) (follow up with in vitro tests) was accepted by the experts on both sides, Prof. Buckton for Richter and Prof. Davies for Schering. Both of these were experts in drug formulation and not medicinal chemists.

The judge at [122] was impressed with the evidence of Prof. Davies to the effect that once the in vitro tests had shown rapid dissolution in stomach conditions, he would not press on with animal trials. He said “there was great force” in that evidence. And he gave a cogent reason, namely pressing on in that way was “strongly contra-indicated by the available evidence.”

Given that “an appellate court should be very cautious in differing from the judge’s evaluation” [of obviousness] (per Lord Hoffmann in Biogen v Medeva [1997] RPC 1 at p.45) it is not surprising that Mr Thorley did not challenge the finding that the inventions were not obvious over Oelkers alone.

Where the judge went wrong, submitted Mr Thorley, was at stages (6) and (7). He wrongly concluded, submitted Mr Thorley (a) that the skilled team would not trouble to read the Krause papers and (b) that if it had, the papers would not have provided sufficient encouragement to try an immediate release formulation in vivo.

As to (a) Mr Thorley had the difficulty that the Krause papers were not concerned with DSP but with SP, mentioning DSP only as its metabolite, were not concerned with contraceptive action, were not concerned with a combination product, gave no details of formulation and so on. The papers were not published in journals read regularly by formulators.

The judge said “there was little to persuade me that a skilled formulator would take the trouble to obtain copies of these papers or review their contents.” There was said to be an error of principle here. That is because the judge considered only a skilled formulator whereas the notional team would have included a medicinal chemist, pharmacologists and biochemists too. So, submitted Mr Thorley, the Krause papers would have been unearthed by a literature search, and considered at the pre-formulation stage of the hypothetical project. Others in the team would have read Krause. He took us through chunks of the cross-examination of Professor Davies to try to establish this. It was all circumlocutory: the direct question: “You may not have read the Krause papers yourself, but others in the team would read them?” was not put.

We are not persuaded that the judge made any error here. Richter called no medicinal chemist to prove that Krause would have been read, yet it had leave to call one. The roundabout way this key part of Richter’s case was sought to be proved suggests that direct persuasive evidence was not available. The judge was entitled to conclude, as he did, that “I am left with no confidence that the results of the literature search would lead the skilled person to read the contents of [Krause]”, [141].

That is an end of the appeal for Mr Thorley had to succeed on both points to win. But we consider Mr Thorley’s second point too. He had a real problem here. The “obvious to try” test for obviousness has significant limitations, as was made plain by Lord Hoffmann in Conor v Angiotech [2008] UKHL 49:

The problem for Mr Thorley was that there was really no or only the slightest expectation of success. He seized on the judge’s finding that a combination of Oelkers and Krause would indicate that “The best that one can say in the light of the data in Krause is that the drug might survive in the final formulation”, [148]. Given that, he submitted that it would be obvious to try to find out if it did by conducting in vivo tests. The judge had, he submitted, taken the “fair expectation” test too literally. And he had overlooked the fact that the avenues for research were so limited.

We are not persuaded by this. It is all too easy to reconstruct with hindsight avenues that might have been pursued. The step from Krause to the invention would have been more of a speculative jump in the dark than anything else. True it is the judge said “might” – but that word in context is not to be read as “likely” or “with a significant prospect of success.” It is closer in meaning to “not impossible.” That does not make out an obviousness attack. This invention was not ob via – in the way.

The secondary evidence, that of how long it took Schering actually to reach the invention, supported the conclusion of invention. The judge said “I think Schering’s history provides a reasonable measure of support for non-obviousness”, [173]. That was entirely the sort of evaluation that a trial judge, having gone into the detail with care, as Floyd J did, is entitled to make.

Finally, there was yet another matter which the judge took into account. It is that even now no one knows why rapid dissolution in the stomach works. The judge said:

This is surely correct: if something is inexplicable even after it is known, it is all the more unlikely to have been predictable (and thus obvious) before.

The upshot is that we conclude that this appeal fails.

I agree.