News
Polymorphs: Clearing the Inventive Step Hurdle at the European Patent Office (EPO)
January 2025
Polymorphic forms of active pharmaceutical ingredients (APIs) can represent commercially valuable compound subject matter and can be critical to providing necessary drug product characteristics and performance. However, patenting polymorphs is not always straightforward at the EPO.
Because polymorph screening is advisable during the early development of a drug candidate, the EPO’s position following T 777/08[1] is that, in the absence of any technical prejudice or unexpected property, the mere provision of a crystalline form of a known active compound cannot be regarded as involving an inventive step. According to T 777/08, when starting from the amorphous form of the API as the closest prior art, there would be an expectation that a crystalline form of the API would have improved filtration and drying characteristics. In these circumstances, “the arbitrary selection of a specific polymorph from a group of equally suitable candidates cannot be viewed as involving an inventive step”[2]. In this article, we explore 5 ways in which the EPO’s inventive step hurdle may be cleared in respect to novel polymorphs. In a nutshell, the chances of meeting the inventive step requirement will be improved if the invention goes beyond what would have been expected from a routine polymorph screen.
Background
Solid forms of active pharmaceutical ingredients (API) can exist in amorphous or crystalline forms. Often, several crystalline forms or “polymorphs” exist, as well as other solid forms of the API such as hydrates and solvates. Hydrates and solvates can also exist in one or more crystalline forms. These polymorphs typically differ in terms of properties, including solubility, dissolution rate, in-vivo bioavailability, hygroscopicity, melting points, crystal habit, and stability (chemical and physical). The identification and characterisation of polymorphs is important because this allows pharmaceutical companies to identify suitable candidates for formulation and dosage form development. Early polymorph screening has become a key part of the drug development process (it is a regulatory requirement), and the patenting of new polymorphic forms can often play an important role in protecting innovative aspects of a drug product.
Why it can be difficult to patent polymorphs
Unfortunately, the patenting of polymorphic forms is not straightforward. Because polymorph screening is advisable during the early development of a drug candidate, the EPO’s established position following T 777/08[1] is that, in the absence of any technical prejudice or unexpected property, the mere provision of a crystalline form of a known API cannot be regarded as involving an inventive step. When starting from the amorphous form of the API as the closest prior art, the EPO believe there would be an expectation that a crystalline form of the API would have improved filtration and drying characteristics. In these circumstances, “the arbitrary selection of a specific polymorph from a group of equally suitable candidates cannot be viewed as involving an inventive step”[2].
Furthermore, as far as polymorphs are concerned, the mere existence of a technical effect may also not be sufficient to demonstrate an inventive step (T 41/17). In this decision, the Board found that the skilled person would generally have screened for the most thermodynamically stable polymorph to allow a formulator to avoid form changes when manufacturing a medicament. Thus, despite the patentee demonstrating that the claimed polymorph was the thermodynamically stable form, the Board held that the identification of such a polymorph would have been expected and its beneficial properties, i.e., not converting to other forms under mechanical stress, were related to the thermodynamic stability of the form. Accordingly, despite the presence of a technical effect, an inventive step was not acknowledged in this particular case.
Notwithstanding these established lines of case law, the EPO frequently grants patents on new polymorphic forms. Below we suggest 5 ways in which inventive step can be demonstrated at the EPO.
Clearing the Inventive Step Hurdle at the EPO
- Demonstrating that a balance or combination of beneficial properties has been achieved
While expected improvements are unlikely to improve the position with respect to inventive step, unexpected improvements should help clear the inventive step hurdle. In two recent decisions, T 2086/21and T 672/21, the Board held that polymorphs having a balance of beneficial properties would not be expected. Accordingly, data showing that the claimed polymorph provides a combination or balance of beneficial properties may improve the inventive step position.
In T 2086/21, the patentee provided evidence that their claimed Form B of apalutamide provided improved hygroscopicity, high thermodynamic stability and high polymorphic stability compared to the physical forms disclosed in the prior art. The established position taken in T 41/17 was that a skilled person would have been expected to seek out the most thermodynamically stable polymorph. However, in T 2086/21, the Board held that even for a polymorph with high stability, i.e. not the most thermodynamically stable polymorph, there was no reason for the skilled person to assume that such a form would also have high polymorphic stability (viewed as a distinct technical effect and not the same as high thermodynamic stability) and display improved hygroscopicity. The beneficial combination of properties, therefore, was deemed sufficiently unexpected to support an inventive step. The Board in this decision also stated that the “try and see” case law cited by the Appellant did not apply since there was no clear pointer in the prior art that would lead the skilled person confronted with the objective technical problem to the solution provided by the claimed form.
Similar conclusions were reached in T 672/21. Here, the patentee had demonstrated that their polymorph of selexipag exhibited intermediate thermodynamic stability, improved industrial processability by way of particle size distribution and improved purity in respect to both residual solvents and residual impurities. Because there was no disclosure or suggestion of this balance of properties in the prior art, this was deemed sufficiently unexpected to support an inventive step.
This is not to say, however, that any combination of effects will be deemed patentable. In a parallel case to T 672/21, the same Board in T 1994/22 found another form of selexipag to lack an inventive step. In the latter case, the patentee’s data showed that the claimed polymorph exhibited the best thermodynamic stability, but only intermediate industrial processability, residual solvent content and residual impurities. In these circumstances, the Board found that there was no balance of beneficial properties, as the claimed form was only superior to the other forms in respect to a single property, namely thermodynamic stability. The fact that this polymorph had intermediate processibility, solvent content and residual impurities was not sufficiently surprising to merit an inventive step.
In summary, therefore, data demonstrating a combination of beneficial properties can help support inventive step. However, the prospects of success will vary on a case-by-case basis, depending on the specific effects demonstrated. The mere inclusion of multiple properties in a data set may not be sufficient to support an inventive step if, in reality, there is merely a single predictable improvement amidst the properties disclosed.
- Demonstrating a departure from routine
It can sometimes be possible to demonstrate an inventive step by showing that an invention marks a departure from routine.
In T 1825/21, the patentee demonstrated an inventive step on the basis that the claimed crystalline form of ceritinib free base could not be prepared by routine steps. The closest prior art was amorphous ceritinib hydrochloride. The claimed invention differed from the closest prior art in two respects: firstly, the ceritinib was crystalline and, secondly, the ceritinib was in free base form rather than a hydrochloride salt.
The Board was of the view that these differences provided the crystalline form with stability and ease of drying. Although, in T 777/08, similar effects were said to be expected from a routine screen for polymorphs, the patentee in T 1825/21 provided evidence that numerous attempts to produce crystalline ceritinib hydrochloride only resulted in amorphous precipitates having chloride levels inconsistent with a stoichiometric salt. Thus, there were clear difficulties in its implementation. Having failed to solve the above problem by providing crystalline ceritinib hydrochloride, the Board found that there would have been no reason for the skilled person to turn to ceritinib free base in the expectation that this would improve stability and ease of drying, especially considering the properties of an API can often be optimised by conversion of the API to a salt form. Accordingly, by demonstrating difficulties with routine steps, and implementing adjustments that were not foreseen by the prior art, the patentee was able to demonstrate an inventive step.
Similarly, in T1422/12, the patentee reformulated the objective technical problem as one that was not considered routine in the context of polymorph screening. In particular, the patentee submitted that the problem underlying the present application was to provide tigecycline in a form more stable with respect to epimerisation. The Board held that, although certain effects were known to be improved by crystallisation, the problem of epimerisation of tetracyclines was not one that was known to be solved by crystallising a drug into a particular polymorphic form. As such, this property was regarded as unexpected and specific to tetracyclines. As the prior art made no link between crystalline forms of tigecycline and epimerisation, the provision of a crystalline form of tigecycline was deemed to be inventive.
- Avoiding overly broad claims
Claims to a single polymorph are often defined by the position of peaks in an X-ray powder diffraction (XRPD) produced using X-rays of a defined wavelength. Typically, 2θ values are given to within a defined error margin. By reducing the number of peaks and/or broadening the error margins, it may be possible to define the polymorph broadly and to limit the chances of third parties establishing non-infringement positions by relying on experimental variability.
In some instances, however, claims that are drafted too broadly may cover more than the intended polymorph. For example, if only a limited number of peaks are listed, mixtures of known polymorphs may exhibit the same peaks. Even if a polymorph is referred to by a particular form, e.g., “Form A”, such a designation is viewed by the EPO as merely a label rather than something that limits the scope of the claim (T 1555/12). Moreover, even if defined as “characteristic peaks”, XRPD peaks are generally considered to define a crystalline form by their presence, not by their strength or intensity. Accordingly, if a claim is merely defined by a small number of peaks, the EPO may take the view that the claim covers more than the intended polymorph and therefore cannot deliver the alleged technical effect across the scope of the claim. In such a scenario, it will be difficult to demonstrate an inventive step, as the claimed subject matter is likely to be considered arbitrary or obvious.
A patentee may be able to improve their position in the above scenario by including fall-back positions that enable the polymorph to be characterised more precisely. This may include e.g., more XRPD peaks or the XRPD peaks may be combined with other characterising information, including Fourier transform infrared spectroscopy (FTIR), Raman spectroscopy, differential scanning calorimetry (DSC),thermogravimetric analysis (TGA), solvent/water content, and/or stability characteristics. It may be advisable to include these additional characterising data in the specification so that you have basis for amending your claim to strengthen your position on inventive step while minimising the risks of any added matter objections.
- Provide additional comparative data (particularly during opposition)
To demonstrate an inventive step, it is usually necessary to show a technical effect over any technical teaching that can be considered as the closest prior art. For example, a patentee may have to compare their new polymorph with the form of API used in a marketed medicinal product. However, in T 559/22, the closest prior art document disclosed two other forms, Forms II and III, in addition to the marketed form, Form I. The Board held that there is no reason to assume that the skilled person would not have realistically started from all polymorphic forms disclosed in the starting document. Because there was no data showing any unexpected effect over Forms II and III, the polymorph was deemed to be an obvious alternative over these forms. Had the proprietor been able to show unexpected technical effects over each of these forms, a different conclusion may have been reached.
- Starting from another crystalline form as closest prior art
Because polymorph screening is often undertaken early during drug development, there is a perception among some EPO practitioners that it can sometimes be more difficult to demonstrate an inventive step over an amorphous API. When starting from crystalline forms of the API, the Boards in T 1684/16 and T 1326/18 found that the fact that a skilled person was taught to investigate polymorphs for the most desirable properties was not necessarily sufficient to render a specific polymorphic with a desired property obvious.
In view of T 1684/16 and T 1326/18, some practitioners have argued that, while a routine polymorph screen might identify polymorphs that, collectively, may be expected to have advantages in terms of filterability or drying when compared to an amorphous form, the possibility of a specific polymorph having a technical effect over another, specific crystalline form is more difficult to predict. Accordingly, once the closest prior art is a crystalline form, some have argued that it may be easier to establish inventive step.
It should be noted, however, that the EPO has not explicitly drawn a distinction between the likelihood of grant when starting from crystalline rather amorphous forms. In fact, in T 1684/16[1] and T 1326/18[2], there were passages in the prior art that taught away from the claimed subject matter. The prior art in these decisions offer an alternative explanation as to why an inventive step was found over known crystalline forms. In fact, when it was argued by the patentee that routine screening might only be routine in the early stages of drug development, the Board in T 559/22[3] held that, while “it may be that an extensive screening for crystalline solid state forms is carried out in the early development stage… there is no apparent reason… why the skilled person should not carry out further screenings at a later stage, at least when faced with the problem of providing a further crystalline solid state form”. Thus, it cannot be said for certain that inventive step is more easily demonstrated when starting from a polymorphic form as the closest prior art. However, it does seem apparent that when several polymorphs exist, demonstrating an improved property of one of the polymorphs as compared to the others to show that the claimed form is not an arbitrary selection from equal alternatives, can help support inventive step by way of an unexpected effect.
[1] Reasons, 4.3.4 of T 1684/16
[2] Reasons, 17.5 of T 1326/18
[3] Reasons, 7.3 of T 559/22
This article was written by Partner and Patent Attorney Hsu Min Chung , Partner and Patent Attorney Balvinder Matharu and Senior Patent Attorney Joanna Pownall
