An infringement theory to be taken with a grain of salt

In a previous post entitled The salt of the judgment, I reported on an order issued in July 2016 and rejecting a request for preliminary injunction against Biogaran’s generic version of the drug rosuvastatin.

The judgment on the merits has now been issued, so this is an opportunity to go back to this interesting case. The problem with this approach is that finding a good title for a blog post in relation with a given case is difficult enough once, and it is even more difficult the second time. By the time I may get to report on a possible appeal ruling, I am afraid I may run short of salty idioms.

So, yes, everything in this case is about salt, or rather salts.

Claim 1 of the patent at stake (EP0521471, owned by the Japanese pharmaceutical company Shionogi, and licensed to the Swedish group AstraZeneca), on which the SPC is based, reads as follows:

The compound (+)-7-[ 4-(4-fluorophenyl)-6-isopropyl-2-(N-methyl-N-methylsulfonylamino)pyrimidin-5-yl)-(3R,5S)dihydroxy-(E)-6-heptenoic acid or a non-toxic pharmaceutically acceptable salt thereof.

The compound in question is rosuvastatin, in its acid or salt form.

The active substance in Biogaran’s generic drug is the zinc salt of rosuvastatin. The zinc salt is not cited in the patent at stake.

As a reminder, the judge in charge of emergency proceedings ruled that the zinc salt of rosuvastatin was in fact excluded from the scope of the patent, when the claim was properly interpreted in view of the description.

Now this preliminary view has been confirmed by the three-judge panel of the Tribunal de grande instance (TGI). As a side note, the judge who issued the interim order (Ms. Courboulay) was not part of this panel.

Here is the relevant part of the judgment (as usual, I took the liberty of cutting excessively long sentences into shorter ones):

[…] [In the patent,] the patentee has mentioned in a clear and precise wording at paragraph [0007] what is meant by “a non-toxic pharmaceutically acceptable salt”, namely “a salt wherein the cation is an alkaline metal ion, an alkaline earth metal ion or an ammonium ion”. [It] does not comprise any indication allowing [the skilled person] to select a salt in another family. [Besides,] the patentee, even if it did not expressly exclude a specific salt, did not, unlike in other paragraphs of the patent, add the expressions “which are not to be considered as limiting”, [0029] […] or “and the like”, [0006] and [0024] […]. [Thus, the skilled person] understands that the salts protected in the patent are defined in a limiting manner as a salt wherein the cation is an alkaline metal ion, an alkaline earth metal ion or an ammonium ion. 

Therefore, and as decided by the judge in charge of emergency proceedings, it derives that the non-toxic pharmaceutically acceptable salts protected by claim 1 are salts wherein the cation is an alkaline metal ion, an alkaline earth metal ion or an ammonium ion.

A very interesting example that claim interpretation in view of the description can be used both ways: not only for broadening the scope of protection of the patent beyond what is literally claimed, but also in some situations for restricting said scope.

As a result of this claim interpretation, the necessary conclusion was that Biogaran’s zinc salt of rosuvastatin did not literally infringe the SCP.

Shionogi and AstraZeneca argued that zinc could be considered as an alkaline earth metal at the filing date. They said that the exact classification of zinc was a matter of debate.

However, the court was not convinced. Based on a 1978 dictionary and on Mendeleev’s table, it appeared that zinc could not be considered as an alkaline earth metal.

Non-toxic salts are useful for other purposes than drugs: salt sculptures for instance.

Another auxiliary argumentation by the plaintiffs was one of contributory infringement: once ingested, the zinc salt of rosuvastatin dissociates, and the anionic active molecule recombines with sodium or hydrogen, to form the sodium salt or the acid form of rosuvastatin. Thus, the zinc salt of rosuvastatin should be considered as an essential means for carrying out the claimed invention.

This bold argument failed – just as it had in the interim proceedings:

[…] The fact that the zinc salt of rosuvastatin, after ingestion, would dissociate in the patient’s body under a neutral acid form may not characterize contributory infringement […]. First, the third party mentioned in article L. 613-4 Code de la propriété intellectuelle [CPI], namely the patient, is not aware of the mechanism of decomposition of the drug within the body, so that he/she does not know under this provision that the ingestion of the drug at stake is intended for implementing the claimed invention. Second, this drug ingestion by the patient is carried out in a private and non-commercial context, so that the patentee’s rights do not extend to such acts. 

The first reason mentioned by the court for rejecting this infringement theory is quite convincing. It is true that the patient has absolutely no idea what the drug becomes after ingestion.

The second reason looks much more challengeable on a legal standpoint. Admittedly, the rights conferred by a patent do not extend to acts carried out in a private and non-commercial context (article L 613-5 CPI). But here, the allegedly infringing act is not the ingestion of the drug by the patient, but the sale of the drug to patients, considered as the supply of essential means for implementing the invention to a third party. And article L. 613-4 CPI clearly states that a party benefiting from the exemption of article L. 613-5 CPI (such as a patient) is not regarded as a party authorized to implement the claimed invention for the purpose of the assessment of contributory infringement.

That said, one good reason was certainly enough to throw out Shionogi and AstraZeneca’s auxiliary infringement theory.

In addition to rejecting the plaintiffs’ infringement claims, the court also had to examine the validity of the EP’471 patent.

The interim order had already stated that there was no serious validity challenge. In the judgment now issued by the three-judge panel, the validity of the patent has been confirmed.

The grounds for nullity which were raised were extension of subject-matter, insufficiency of disclosure and lack of inventive step.

Regarding extension of subject-matter, original claim 1 in the application as filed was directed to a compound defined by a Markush formula with five variable groups X, R1, R2, R3 and R4.

Two examples of the application disclosed calcium rosuvastatin and sodium rosuvastatin.

The defendant argued that the acid form of rosuvastatin, on the other hand, was not directly and unambiguously disclosed in the application as originally filed.

But the court was satisfied that claim 1 of the patent was obtainable by the skilled person based on the examples of calcium and sodium rosuvastatin, and on a paragraph of the description which provided a broader definition for group R4 in the general formula, namely: hydrogen (corresponding to an acid form of the molecule), or a lower alkyl group, or a cation capable of forming a non-toxic, pharmaceutically acceptable salt.

What is not explicitly discussed in the court’s reasoning is the fact that the definition of group R4 was restricted to arrive at claim 1 as granted, since the option of a lower alkyl group is no longer covered in the patent.

Turning now to sufficiency of disclosure, Biogaran raised two objections:

  • there is no information in the patent how to select or use salts of rosuvastatin other than the exemplified sodium and calcium salts;
  • the patent does not sufficiently establish the existence of a therapeutic effect, as it only disclosed an in vitro test, and did not contain any precise indication on a possible dosage regimen.

Regarding the first objection, it is certainly relevant that only certain types of salts were held by the court to be covered by claim 1, namely alkaline metal, alkaline earth metal and ammonium salts. Therefore, it was immaterial for the appraisal of sufficiency that the zinc salt for instance is not discussed in the patent, since this salt is not covered by claim 1.

The court noted that the patent discloses a process for making various salts, and that Biogaran did not demonstrate any difficulty for carrying out the invention with the different forms of rosuvastatin that are claimed.

As for the second objection, the court was satisfied that, since the patent relates to new molecules, an assay comparing the biological activity (namely the activity of inhibition of HMG-CoA reductase) of rosuvastatin to that of the prior art drug mevinoline, is a sufficient disclosure of the therapeutic effect.

Lastly, inventive step.

Biogaran’s argument was based on two prior art documents, Bayer and Beck, which both disclose a broad family of compounds having an HMG-CoA reductase inhibitory activity. Biogaran argued that rosuvastatin can be obtained by way of a selection from the teaching of these documents, and that this selection is arbitrary, since the therapeutic effect of rosuvastatin is the same as that of the other compounds of the broad family explicitly taught in Bayer and Beck.

The court held that this reasoning is based on hindsight and thus impermissible. The court stated in particular that the skilled person would focus on options described as preferred in the prior art, and would thus not study all possible compounds at random. In this case, there was no pointer in the prior art leading the skilled person to rosuvastatin.


CASE REFERENCE: Tribunal de grande instance de Paris, 3ème chambre 2ème section, February 9, 2018, Shionogi Seiyaku Kabushiki Kaisha, AstraZeneca & AstraZeneca UK Ltd v. Biogaran, RG No. 16/13292.

Patentability in small doses

Dosage regimen inventions are one of those subjects wherein French law tends to be unique in the European patent law landscape.

By way of a reminder, patent eligibility of claims directed to dosage regimen inventions was denied in a number of court decisions. In the finasteride litigation, the Paris Cour d’appel finally seemed to align with the principles set out in decision G2/08 of the EPO’s Enlarged Board of Appeal, as it acknowledged that posology features are allowed in further medical use claims. See a summary in a previous post here.

However, first instance judges did not seem willing to follow this case law, as shown by two further decisions from the Paris tribunal de grande instance (TGI) reported on at the beginning of this other post.

In December 2017, the Cour de cassation issued its ruling on the finasteride case. The main topic of the decision however is sufficiency of disclosure, as explained in this post. Some have argued that this decision from the supreme court implicitly “acknowledged the patentability of dosage regime claims“. I tend to disagree, as it seems to me that this was simply not decided upon by the cassation judges.

A few months later, we now have a clear confirmation that the issue of patent eligibility of dosage regimen inventions is absolutely not settled.

The case at hand relates to European patent No. EP 1448207 owned by Hungarian company Richter Gedeon Vegyeszeti Gyar RT (Richter) and licensed to French company Laboratoire HRA Pharma (HRA). In June 2013, the generic drug manufacturer Mylan initiated nullity proceedings in front of the Paris TGI. Mylan launched a generic drug in early 2014. Richter tried to get a preliminary injunction against Mylan, which was denied in June 2014.

The French part of the European patent was voluntarily limited in front of the Institut National de la Propriété Industrielle in October 2014, by turning product claim 1 into an EPC 2000-type therapeutic use claim; and by merging Swiss-type claims 2 and 3 together.

On June 9, 2015, the TGI handed down its decision on the merits, revoking the patent. Richter and HRA appealed.

On March 2, 2018, the Cour d’appel confirmed the first instance decision.

The drug at stake in this lawsuit is levonorgestrel, a hormonal substance used as an emergency birth control medicine.

Claim 1 of the patent as limited reads as follows:

Pharmaceutical composition as single application dose, containing 1.5 ± 0.2 mg of levonorgestrel as active ingredient in admixture with known excipients, diluents, flavoring or aromatising agents, stabilizers, as well as formulation-promoting or formulation-providing additives, commonly used in the pharmaceutical practice, for use in emergency contraception by administering a single application dose up to 72 hours after the coitus.

Claim 2 of the patent, also modified by way of the national limitation, is the following:

Use of 1.5 ± 0.2 mg levonorgestrel for the preparation of a pharmaceutical for emergency contraception by administration of a single application dose up to 72 hours after the coitus.

At the priority date of the patent, levonorgestrel was already widely known and used for emergency contraception. It was administered in two doses of 0.75 mg each. The invention thus consisted in replacing these two doses by a single dose of 1.5 mg.

Dosage regimen patents in France are like a game of Hide & Seek.

The court noted the following:

The invention neither modifies the sought contraceptive purpose, nor the used substance (levonorgestrel), nor the total dose of 1.5 mg, nor the administration of the product within 72 hours of non-protected coitus. 

Indeed, it is not challenged that taking two doses of 0.75 mg levonorgestrel is tantamount to one dose of 1.5 mg levonorgestrel, as the additives are not the subject-matter of the invention and anyway are identical for Noverlo 1.5 mg and for Noverlo 0.75 mg.

The court then further held:

Besides, the description of the patent does not contend that the invention makes it possible to obtain better results to avoid pregnancies, or to reduce side effects, but that its purpose is to solve the problem of the difficulty for patients to comply with the taking of the second dose within a period of twelve hours from the first one, while achieving at least the same results without additional side effects. 

Thus, and insofar as the sole contribution of the ‘207 patent consists in taking the product which is identical in its substance, in its total dosage and for the same indication, in one take instead of two without any novel technical contribution or benefit other than the comfort of a single take, the first instance court rightly held that the invention is not patentable under article 53(c) of the European patent convention. 

There you have it, dosage regimen inventions can still be held non-patentable as relating to mere methods of treatment.

This applied similarly to purpose-limited composition claim 1 and to Swiss-type claim 2. This is not surprising as French courts do not typically attach importance to the exact manner in which claims are drafted. They focus on what the invention really is about.

As a further comment, we should never read too much in any given court decision.

In particular, I do not believe the present decision to be a complete reversal relative to MSD v. Actavis. In the present ruling, the court insisted on the fact that there was absolutely no technical contribution, in their opinion: same total dosage, no reduction in side effects, no efficacy improvement. In a different context, with a new dosage providing for instance improved efficacy or fewer side effects, the court might have come to a different conclusion, based on the existence of an actual technical contribution to the art.

As if one deadly wound were not enough, the court inflicted two additional ones to the patent, in a clear effort to make Richter’s way to a cassation appeal as difficult as possible.

The court thus held that the claims of EP’207 lacked novelty over clinical trials on the 1.5 mg dosage which were publicly reported on by the World Health Organization before the priority date.

And the court finally held that the claims lacked inventive step, also in view of the clinical trial reports.

As a final remark, it is somewhat paradoxical that the patent was revoked as relating to a method of treatment, whereas a method of contraception is in principle not considered at least by the EPO as a method of treatment – as pregnancy is not a disease. See example 3 in section G-VI, 7.1.2 of the Guidelines for examination. I do not know whether the argument was raised during litigation or not.

That said, if one adopts this approach, it then means that perhaps the EPO should not have granted claim 2 of the patent at least in this specific form, as the exceptional Swiss-type claim drafting format is not applicable to non-medical uses (so that the claim would or should have been found to lack novelty).


CASE REFERENCE: Cour d’appel de Paris, pôle 5 chambre 2, March 2, 2018, Richter Gedeon Vegyeszetu Gyar RT & SAS Laboratoire HRA Pharma v. SAS Mylan, RG No. 15/16651.

New opposition extensions

Since July 1, 2016, the EPO has implemented a new procedure to accelerate opposition proceedings. The EPO’s target is to reduce the total time needed for a decision in “straightforward cases” to 15 months, calculated as from expiry of the opposition period.

In keeping with this ambitious goal, the time period allocated to patent proprietors to reply to an opposition has been reduced. Or, to be more precise, the nominal period remains four months from the issuance of a communication by the EPO. But, formerly, a two-month extension was automatically granted upon request, and my understanding is that use was very often made of this possibility.

Nowadays, this extension is only “granted in exceptional, duly substantiated cases“, as indicated in the Guidelines, E-VIII, 1.6.

A few weeks ago, I filed a substantiated request for such an extension of the time limit on behalf of a patent proprietor, which was denied. I must say that I felt a little bit upset (if not offended), especially because I had the impression that in some occasions, when the shoe was in the other foot (i.e. I was representing the opponent), similar requests had been granted to the adverse patent proprietor.

Thinking a little bit about this, I realized that we have little to no information as to what an “exceptional” case is supposed to be for the EPO.

The only further, indirect, indication in the Guidelines is the paragraph which follows the one quoted above, and which concerns “other proceedings, for any communication raising a matter of substance“, whenever an extension up to a total of more than six months is requested. It is specified that this type of extension “should be allowed only exceptionally, when the reasons given are sufficient to show convincingly that a reply in the period previously laid down will not be possible. Such exceptional circumstances might be e.g. the fact that a representative or client is so seriously ill that he cannot deal with the case in time; or the need to perform extensive biological experiments or tests. On the other hand, foreseeable or avoidable circumstances (e.g. leave, pressure of other work) should not be accepted as a sufficiently exceptional circumstance”. 

Whether the same standard applies to a two-month extension in opposition proceedings (only in an “exceptional case“) as to an extension of more than two months in examination proceedings (which should be “allowed only exceptionally“) is anyone’s guess. I assume that formality officers have more specific internal guidelines, but by definition we know nothing about those.

In order to find out more, and without having the time to perform a very thorough analysis, I decided to take a random sample of opposition cases.

More precisely, I had a quick look at a number of opposed patents granted in June 2016 (opposition time limit in March 2017). I discarded all unusual cases (opposition withdrawn early, request for revocation of the patent filed in response to the opposition). There were approximately 150 cases remaining in my sample.

The first teaching is that no request for extension of time is filed in the vast majority of cases.

In my sample, the rate of request for extension of time was less than 18%. I assume that it may be even lower nowadays, as professionals are probably even more aware of the new policy now than they were back in March or April 2017. This is probably a significant change in the behavior of patent proprietors.

The second teaching relates to the success rate of requests for extension. I found that an extension is refused approximately 60% of the time.

Conversely, it is directly granted in approximately 30% of cases. The remaining cases (approximately 10%) are those where a request for extension is initially denied and then granted further to a second attempt.

Extension granted.

It is then interesting to enumerate various reasons offered in requests for extension, whether they were deemed sufficient or not by the EPO.  Please note that some requests contained several of the reasons listed below.

Some of the losing reasons first:

  • No justification at all – a non-starter of course.
  • Complexity of the opposition proceedings.
  • Necessity to coordinate the opposition case with other cases (parallel divisional applications, national invalidation trials).
  • A large number of cited documents, requiring extensive discussions.
  • A large number of pages to translate.
  • Difficulties understanding the opponent’s case.
  • Summer break season making a number of people unavailable for comments.
  • Instructions from a foreign client awaited.
  • Difficulties getting in touch with the inventor.
  • Difficulties getting in touch with staff having left the company.
  • Waiting for input from R&D.
  • Intention to prepare affidavits.
  • Change of representation.

And now some of the winning reasons:

  • Technical expert had a car accident.
  • Several oppositions filed.
  • Public prior use argument raised by the opponent.
  • Experiments being carried out by the proprietor.
  • Need to study an experimental report filed by the opponent.
  • Difficulties getting in touch with the inventor.
  • Many lengthy prior art documents cited.
  • Change of representation.

As a takeaway, I would say that, in order for a request for extension of time to be considered as relating to an “exceptional case“, any reason related to the normal conduct of business should be avoided.

Holidays, the need to coordinate several cases or to receive instructions from a distant client are not sufficient by themselves.

On the other hand, a truly unexpected event, such as a car accident, is not necessarily required.

Opposition proceedings that are more complex than usual often also qualify. But it is then necessary to be specific as to the unusual complexity of the case. Experimental testing, public prior use and multiple oppositions apparently meet the standard.

That said, there is still a grey area.

For instance, in a number of cases, it seems that a change of representation (which is very common after an opposition is received) is by itself sufficient to guarantee an extension of time – but not always. Similarly, difficulties to reach the inventors, or the number and length of cited documents, are sometimes viewed as giving the right to an extension of time, and sometimes not.

There was even a real surprise among the selection of cases that I looked at. Namely, a law firm explained that they had difficulties getting instructions from their client in Texas, as there had been disruptive floods in this area. Apparently, this was not deemed exceptional enough for the formalities officer…

In summary, you can never know for sure whether your request for extension of time will be granted or not, but looking at examples in other cases certainly helps making a prediction and tailoring your arguments.

On the other hand, as long as the patentee files at least a couple of requests in due time (typically, that the opposition be rejected, and that oral proceedings be summoned prior to any adverse decision), I think that in many cases it is still OK to submit a detailed argumentation shortly afterwards. There is no basis in the Guidelines to discard the patentee’s argumentation simply because it was filed after the initial deadline. In fact, arguments and amended claims filed before the oral proceedings, as long as within the R. 116 time limit, are usually always admitted into the proceedings. Only the filing of evidence within the R. 116 time limit is typically challenged and examined as to admissibility.