Polymorphs are different crystalline forms of the same chemical compound. Elemental carbon exhibits polymorphism. Both graphite and diamond are pure carbon but have different crystal structures resulting in very different properties (and price tags).
Polymorphism of drug molecules is an important consideration in the pharmaceutical space as different polymorphs can lead to improved properties such as improved solubility, dryability and filterability. Moreover, many drugs will only receive regulatory approval for a single polymorph. Consequently, obtaining a patent to a specific polymorph of a pharmaceutical product can provide a critical layer of protection for drug candidates.
We are often asked if polymorphs can be patented at the EPO and, while there is no specific exclusion from patentability for polymorph claims, it is not always easy for these to meet the EPC's requirements with regard to inventive step.
The EPO's key "polymorph inventive step" decision is T 777/08. In this decision, the Board held that the claimed polymorphic form was not inventive, since the advantages of the claimed polymorph (improved dryability and filterability) were predictable for the skilled person, and the skilled person was motivated to look for the existence of polymorphic forms. The Board further noted that, in this case, the Patentee had arbitrarily selected a specific polymorph from a range of equally suitable candidates.
Following T 777/08, EPO case law has developed which indicates that polymorphs may be inventive, if the specific polymorph has unexpected advantages. For example, in T 1422/12 the claimed crystalline form of tigecycline had improved stability with respect to epimerisation, resulting in improved biological activity. Since this improvement in biological activity could not have been predicted, an inventive step was acknowledged.
In both T 1555/12 and T 0517/14, polymorph claims were found to be allowable as the claimed crystalline forms were found to have improved stability in view of other polymorphic forms. Thus, the claimed polymorphic forms were not "arbitrary" selections, and T 777/08 did not apply. T 2114/13 reached a similar conclusion. In paragraph 5.7.4 of that decision, the Board specifically commented on T 777/08, noting that "In that decision, the specific polymorph claimed was found to be an arbitrary choice from equally suitable candidates [...] In the present case, the board has no doubt that not all crystalline forms of febuxostat are equally suitable candidates to solve the problem of providing a crystalline form with improved polymorphic stability. Indeed, the patent in suit also mentions crystalline forms, which are not polymorphically stable (see paragraphs [0037] to [0038])."
The most recent decision in this area is T 0041/17 (issued earlier this year). This decision highlights that, in addition to an unexpected technical effect, the choice of closest prior art may be crucial.
In T 0041/17 the claims at issue related to polymorph I of sorafenib tosylate. The Patentee argued that polymorph I had improved resistance to mechanical stress and was able to provide data to show that samples of polymorph I did not degrade under mechanical stress, but that polymorphs II and III did. Following T 0517/14 and T 2114/13 the Patentee argued that their claims were patentable, since this improved stability could not have been predicted. While this argument was successful in front of the Opposition Division, the Board of Appeal disagreed. The closest prior art - agreed by both parties - disclosed sorafenib tosylate in an unspecified form. The Board agreed with the Appellant (Opponent) that the skilled person was motivated to find the most thermodynamically stable crystalline form, and that this would inevitably be resistant to mechanical stress. Thus, the claims lacked an inventive step following T 777/08.
The key difference between decisions T 0517/14 and T 2114/13 and T 0041/17 lies in the disclosure of the closest prior art.
In both T 0517/14 and T 2114/13, the closest prior art already disclosed polymorphic forms. Thus, the invention lay in the finding of a further, more stable polymorph. In contrast, in T 0041/17, the closest prior art disclosed an unspecified form of sorafenib tosylate and the invention only related to finding the most stable polymorph form.
The crux here is that, when a polymorphic form exists in the prior art (and is identified as being stable), there is no motivation for the skilled person to find alternative polymorphic forms or to have any reasonable expectation of finding a better one. In contrast, when no polymorphs are known, the skilled person is inevitably motivated to find the "best" one.
The choice of the closest prior art document is always critical in determining an inventive step at the EPO. In the specific case of polymorph claims, it may work in the Applicants/ Patentees favor to find a prior art document that discloses at least a crystalline form of the compound in question
Key takeaways:
- Claims to polymorphs can be patentable at the EPO if they have unexpected properties;
- Evidence of polymorphs that are disadvantageous is likely to be helpful; and
- Data is crucial to support points (1) and (2) above.
In this article, we have specifically limited our discussion to the EPO's consideration of inventive step. However, polymorph claims can be vulnerable to lack of clarity and/or lack of sufficiency rejections at the EPO.
Originally published June 9, 2020
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
