Nice models required

The title of this post was not – only – selected in order to artificially boost the stats of this blog, although I guess this could be a possible side effect. No, there is an actual relationship with today’s discussion, which is again the appraisal of sufficiency of disclosure for some pharmaceutical patents.

Very often, these patents are filed too early in a drug development process to comprise any clinical data. At best, they only contain preliminary results from in vitro or animal sudies. Whether this is deemed to be sufficient for the skilled person to implement the claimed invention will depend on the models that are used – that is, the disease models, I am afraid.

I will now leave the floor to Lionel Vial who will explain how this all works.

Renaud’s previous post related to the French way of dealing with the sufficiency of disclosure requirement regarding therapeutic purpose-limited product claims (compound X for use the treatment of Y) when there is a doubt that the therapeutic technical effect has been attained. But how does the EPO deal with it nowadays?

Decision T 2059/13 of December 7, 2015 was rendered on the appeal lodged by patentee Otsuka Pharmaceutical against the decision of the opposition division to revoke European patent No. 1712225.

The patent was revoked under the ground of Article 100(b) EPC (i.e. insufficiency of disclosure) by applying the landmark decision T 609/02 of October 27, 2004, which was discussed in Renaud’s previous post. As a reminder, the catchword of this decision is that

If the description of a patent specification provides no more than a vague indication of a possible medical use for a chemical compound yet to be identified, later more detailed evidence cannot be used to remedy the fundamental insufficiency of disclosure of such subject-matter (emphasis added).

Claim 1 of the main request in the appeal proceedings at hand read:

A compound which is a pharmaceutically acceptable acid-addition salt or solvate of a carbostyril compound of the formula (1):

Formula

wherein the dotted line represents a single or a double bond, for use in the treatment of disorders of the central nervous system associated with 5-HT1A receptor subtype, selected from

(i) bipolar I disorder with most recent hypomanic, manic, mixed, depressed or unspecific episode, and

(ii) bipolar II disorder with recurrent major depressive episodes with hypomanic episodes, and cyclothymic disorder.

The other claims also related to the compounds of formula (1) for a further medical use and it was not contested by the parties that in such cases, for the requirement of sufficiency of disclosure to be fulfilled, the suitability of these compounds for the claimed therapeutic application must be disclosed (point 4.1 of the Reasons).

The patentee argued that the facts and circumstances of the present case differed from those underlying T 609/02 in which the chemical structure of the compounds was not identified (point 4.2.2 of the Reasons). Indeed, the above claims are restricted to a group of only two, well identified, compounds.

The Board only partly concurred with the patentee and explained that the usefulness of case law is not confined to similar or identical facts, but lies in the principles or guidance which can be extracted from earlier cases (point 4.2.3 of the Reasons).

In accordance with these considerations the Board then offered the following statement:

Therefore, for a patent claiming a compound for use in therapy, grounds under Article 100(b) EPC will prejudice the maintenance if the application does not disclose the suitability of the product for the claimed therapeutic application to the skilled person using its common general knowledge. Only once this evidence is available from the patent application, may postpublished evidence be taken into account when assessing sufficiency of disclosure (point 4.2.4 of the Reasons, emphasis added).

Going back to the particulars of the case at hand, the patent in suit properly disclosed and proved that the claimed compounds bind to a receptor called 5-HT1A, or in other terms were 5-HT1A agonists – this was not challenged by the respondents. But the real issue was the link between this biochemical property and the treatment of bipolar disorders.

In this respect, the Board found that the patent as disclosed at its filing date did not render the suitability of either of the compounds of formula (1) for the treatment of any type of bipolar disorder plausible; nor did it provide the information that there is a clear relationship between 5-HT1A receptor agonism and the suitability for the treatment of bipolar disorder (point 4.4.5 of the Reasons).

The Board went on to consider that there was no evidence on file showing that the person skilled in the art was in the possession of common general knowledge at the filing date of the patent in suit (only represented by basic handbooks and textbooks on the subject in question, see points 4.5.1 and 2 of the Reasons) which, together with the disclosure of the application as filed, led to the direct and unambiguous conclusion that 5-HT1A agonists in general, or either of the compounds of formula (1) in particular, were useful in the treatment of any type of bipolar disorder (point 4.5.3 of the Reasons).

Eventually, the Board concluded that the application as filed in combination with common knowledge at the filing date did not disclose the suitability of either of the compounds of formula (1) in the treatment of any type of bipolar disorder. Consequently, the minimum requirements set out in T 609/02 for taking into account post-published evidence were not met (point 4.5.3 of the Reasons).

Modelling of bipolar disorders at a very early stage.
Modelling of bipolar disorders at a very early stage.

The present decision follows the jurisprudence set by T 609/02 (point 9 of the Reasons) and confirmed inter alia by T 433/05 of June 14, 2007 (point 28 of the Reasons), T 801/06 of March 4, 2009 (point 25 of the Reasons), T 1437/07 of October 26, 2009 (point 37 of the Reasons), T 866/08 of September 16, 2010 (point 2 of the Reasons) (kindly brought to our attention by our reader Raoul), T 1685/10 of June 6, 2011 (point 3.1 of the Reasons), and T 801/10 of July 8, 2014 (point 4.1 of the Reasons).

However, it adds a twist to the existing case law by requiring that the skilled person can only rely on common general knowledge, represented by basic handbooks and textbooks, and excluding patent literature and scientific articles, to determine if the experimental data presented in a patent is representative of a metabolic mechanism specifically involved in the disease purported to be treated according to the claimed invention.

This was not expressly mentioned in T 609/02, which although it stated that:

As a consequence, under Article 83 EPC, unless this is already known to the skilled person at the priority date, the application must disclose the suitability of the product to be manufactured for the claimed therapeutic application (point 9 of the Reasons for the Decision, emphasis added),

also considered that:

It is required that the patent provides some information in the form of, for example, experimental tests, to the avail that the claimed compound has a direct effect on a metabolic mechanism specifically involved in the disease, this mechanism being either known from the prior art [i.e. not only from the common general knowledge of the skilled person] or demonstrated in the patent per se. Showing a pharmaceutical effect in vitro may be sufficient if for the skilled person this observed effect directly and unambiguously reflects such a therapeutic application […] (point 9 of the Reasons for the Decision, emphasis added).

In fact, it is not clear if the skilled person mentioned in T 609/02 should be akin to the one of Article 100(a) EPC, who has access to all the prior art, or to the one of Article 100(b) or of Article 100(c) EPC, who has only access to the contents of the patent and common technical knowledge.

The present decision appears to have decided for the latter solution, while previous decisions did not seem to set restrictions on the type of prior art that should be relied on for assessing the suitability of a product for a claimed therapeutic application (see for example T 433/05, point 29 of the Reasons and T 801/06, point 29 of the Reasons – even though in the latter case the prior art documents used were cited in the opposed patent, but this was not mentioned by the Board).

We cannot foresee if this decision will set a new trend in applying the teachings of T 609/02, but it is surely advisable for applicants to strengthen the description of the in vitro or in vivo disease models relied on in applications containing therapeutic purpose-limited claims, in particular by fully citing the scientific literature on which such models are based.

We would also like to add that even though this decision seems to add a further burden on applicants in regard of the sufficiency of disclosure requirement, it may conversely be a benefit to them when considering the novelty requirement. Indeed, as is clearly expressed in the above-mentioned decision T 1437/07:

A disclosure in a prior art document is novelty-destroying only if the teaching it contains is reproducible. This need for an enabling disclosure is in conformity with the principle expressed in Article 83 EPC. Thus, the requirements of sufficiency of disclosure are identical for a prior art document and a patent (point 25 of the Reasons, emphasis added).

Thanks Lionel! I guess the bottom line is that EPO case law is still in an adjustment phase as to the appropriate sufficiency threshold for second medical use patents. Although this topic is of crucial importance for all practitioners in the pharma industry, I think it is rather unlikely to find its way up to the Enlarged Board of Appeal, which – for better or worse – less often deals with substantive issues of patentability than with procedural questions.


CASE REFERENCE: Board of Appeal 3.3.01, T 2059/13, Otsuka Pharmaceutical Co. Ltd. v. Stada Arzneimittel AG et al., December 7, 2015

You only die twice

Some readers may have a sense of déjà vu when reading today’s post, as I am going to talk once again about the finasteride saga.

In Lionel Vial’s first post and second post on the topic, it was explained how the Cour d’appel de Paris confirmed the revocation of the French part of Merck’s patent No. EP 0724444 (EP’444) for lack of novelty. The case was remarkable in that the court endorsed the principles set out by the Enlarged Board of Appeal in G 2/08: they held that second medical use claims containing a dosage regimen feature are not excluded from patentability and that novelty can be conferred by the dosage regimen feature. Yet, the court also held that, in this case, the claimed invention was in fact not novel over the prior art. This ruling was the result of a revocation claim brought by Actavis.

But there was another case in parallel to this one, with a revocation claim brought by Teva. This led to another ruling on the same date, where the same panel of the Cour d’appel confirmed again that the French part of EP’444 was invalid – but on a different ground, namely insufficiency of disclosure. So, I should probably say a word on this one as well, not only because Teva was represented by my former colleagues of August & Debouzy, but also because the court’s reasoning is in my view as interesting as in the Actavis case.

Readers may remember that the main claim of EP’444 was the following:

The use of 17β-(N-tert-butylcarbamoyl)-4-aza-5- alpha-androst-1-ene-3-one for the preparation of a medicament for oral administration useful for the treatment of androgenic alopecia in a person and wherein the dosage amount is about 0.05 to 1.0 mg.

Here are now Merck’s explanations regarding the nature of their invention, in the court’s own words:

They state that the technical problem to be solved was to identify a compound the mode of administration of which would offer the best guarantee in terms of safety (low dosage) and efficiency, so as to provide an improved treatment of androgenic alopecia. They explain that they have very surprisingly discovered that the oral administration of low daily doses of the finasteride compound, from 0.05 mg to 1.0 mg, was particularly effective for treating this condition.

The court examined whether the description of the patent was sufficiently clear and complete to enable the skilled person to carry out this invention. In this respect, they relied on the following standard:

Regarding an invention relating to a further treatment, even if, as rightly put by the appellant, it is not necessary to demonstrate clinically the therapeutic effect, nevertheless the pharmaceutical effect demonstrated in the application must directly and unambiguously reflect the claimed therapeutic applications, so that the skilled person understands based on commonly accepted models that the results reflect these therapeutic applications.

[…]

The inventor must indicate that the result was investigated and exists, by any experimental or non-experimental information, which establishes and makes explicit the claimed pharmaceutical effect, as from the filing date. 

In other terms, second medical use inventions are subjected to a somewhat stricter sufficiency test that other inventions. The basic test for most inventions is whether it is possible to make the claimed product or implement the claimed use without undue burden. The stricter test applied here demands in addition that evidence of the pharmaceutical effect be present in the application itself. Data form clinical tests is not required, as those are usually conducted quite late in the drug development process. But some kind of research must at least have been conducted and some results obtained, that the skilled person can interpret as reflecting the claimed therapeutic use.

This seems like a sound approach in order to curb speculative patents, or “paper” patents, which could be used as an evergreening weapon, so as to preempt mere ideas or working hypotheses and block all future R&D efforts.

To some extent, I think the Boards of appeal of the EPO have adopted a similar principle, notably in T 609/02. An entire section of the patent attorney bible Case Law of the Boards of Appeal of the European Patent Office, 7th edition, September 2013 (II.C.6.2) is precisely dedicated to the particular “level of disclosure” which is required for medical use claims.

This is not to say that the principle is applied with the same strictness in Munich as in Paris. Practically speaking, the heightened sufficiency test mainly seems to be used by the Boards to strike out claims seeking to cover excessively large classes of compounds for certain uses, notably classes of compounds defined according to their purported function or use (also known as reach-through claims), as well as medical use claims wherein there is no demonstration of a direct link between the claimed compound(s) and the medical treatment at stake; but patents such as the present one, directed to a new dosage in an already known medical use of a given compound, are probably much more immune to an insufficiency challenge at the EPO.

Anyway, the Cour d’appel was clearly dissatisfied with the contents of the EP’444 patent:

[…] The description does not indicate what is the advantage or the technical effect resulting from this type of oral adminsitration. It does not contain any element showing the potential efficacy of the lower finasteride dosage, so that this mode of administration has no relevance for the skilled person.

The purpose of the invention according to the description is to reduce the amount of administered finasteride relative to the acceptable dosage already known from the prior art for an indication already disclosed, but the description does not comprise any information on the novel effect of the claimed posology and the particular properties of this new therapeutic application. […]

Moreover, no relevant and convincing result is provided in order to justify the claimed and yet undescribed pharmaceutical effect. 

The court turned in particular to the example section in the EP’444 patent and held that none of the examples was relevant to the claimed therapeutic use.

I had a look at this example section myself, and it seems that the court made a good point here. Examples 1 and 2 relate to methods for making finasteride. They are thus irrelevant to the claimed therapeutic use. Example 3 discloses protocols for the preparation and dosage of 5α-reductase – but no results.

Example 4 describes a protocol for measuring haircount in subjects. Its conclusion is rather vague, though: “Using the above-described methodology, it can be shown that administration of finasteride, in dosages per day per patient of, for example, 1 mg/day or 0.2 mg/day, are useful in the treatment of androgenic alopecia, and promote hair growth in patients with this condition“. This is still largely unspecific in terms of technical effect(s) achieved and has the air of a prophetic example.

Finally, example 5 is sufficiently short to be reproduced here in its entirety: “In another test, finasteride was orally administered for 6 weeks to men with male pattern baldness at doses of 0.2 mg/day and 1.0 mg/day (and, for comparison, 5.0 mgs/day). The results of this test showed a significant reduction in DHT content in scalp tissue of the test participants“. Again, this is largely unspecific in terms of the improvement in the treatment of patients that can be achieved using to the claimed dosage of 0.05 to 1 mg finasteride.

As a line of defense, Merck provided external evidence such as an expert opinion and results from a study conducted in 1993 – the same year the priority application was filed. The court paid little attention to this external evidence and simply stated that it is of no relevance.

This is of course a crucial point. How much you allow a patent proprietor to rely on post-published evidence of a technical effect can make a world of a difference in terms of validity in the pharma and biotech field.

The court’s final conclusion was worded as follows:

[…] The skilled person is unable to reproduce the invention, since he is kept unaware of any specific technical teaching, and he must therefore perform a research program on his own. 

Another finasteride post means another hair picture - this time Picasso-style.
Another finasteride post means another hair picture – this time Picasso-style.

As to the obvious question, why rely on different invalidation grounds in the two parallel nullity suits for knocking the same patent out, I assume that the answer is that courts in this country are very strictly bound by the submissions of the parties, and the plaintiff’s submissions were probably different in the respective cases.

Anyway this striking example of overkill does tell us something about our judges’ approach to what they may perceive as “weak” patents. Sometimes we get the impression that once a court is convinced that the invention at stake does not bring a contribution to the art which is specific and significant enough – and is therefore unworthy of patent monopoly – the ground on which they decide to invalidate the patent is not that important.

So, in this case, lack of novelty was as good as insufficiency of disclosure, given the absence of a “specific technical teaching” (as stated in the Teva judgment) or of a “different technical teaching” (as stated in the Actavis judgment). I am pretty sure that if there had been yet a third parallel case, the patent could also have been invalidated for lack of inventive step in view of the same findings.

As a last word on this finasteride saga (until a report on a potential cassation appeal?), the court noted in the decision that the EP’444 was found valid in the UK, Germany, Italy and the Netherlands. Oops! So much for pan-European consistency.

On the other hand, based on the slim contents of EP’444, I have sympathy for the position of our local judges. Merck’s invention may have been a real and valuable one – this is very possible. But one cannot really tell so based on the patent itself.


CASE REFERENCE: Cour d’appel de Paris, Pole 5, 2ème chambre, January 30, 2015, Merck Sharp & Dohme Corp. v. Teva Santé et al., RG No. 10/23603.