The link to the decision is provided after the jump, as well as (an extract of) the decision text.
2. The respondent put forward a novelty objection to the subject-matter of claim 1 based on D1.
3. D1 addresses the problem of providing crystalline 2'-FL. It states (lines 1 to 6) that purification of 2'-FL by repeated chromatography or via its tosylhydrazone does not immediately yield crystalline 2'-FL. Rather, a syrup or, after treatment with alcohol, an amorphous white powder is obtained first.
3.1 According to D1 (lines 6 to 10), crystalline 2'-FL can be obtained from the above-mentioned syrupy 2'-FL as follows (denoted as "method 1" by the parties):
"Nach längerem Aufbewahren von 2.1 g eines mit Wasser abgedampften Sirups bei ~20° schieden sich am oberen Teil der Kolbenwand einige weiße Kriställchen ab, mit deren Hilfe sich der gesamte Sirup über Nacht in ein lockeres Kristallpulver verwandeln ließ."
[translation by the board: After storing 2.1 g of a water-evaporated syrup for an extended period of time at approximately 20°, some white crystals separated at the upper part of the wall of the flask, with the help of which the entire syrup could be transformed into a loose crystalline powder overnight.]
3.2 D1 further discloses (lines 10 to 17) that, starting from amorphous 2'-FL, crystalline 2'-FL can be obtained as follows (denoted as "method 2" by the parties):
"Ein anderer Ansatz, in dem nach mehrwöchigem Aufbewahren bei +4° spontan Kristallisation eintrat, war durch Lösen von amorphem Trisaccharid in wasserhaltigem Methanol und Zugabe des gleichen Volumens n-Butanol sowie einiger Tropfen n-Hexylalkohol bereitet."
[translation by the board: Another approach, in which spontaneous crystallisation occurred after several weeks of storage at +4°, was conducted by dissolving amorphous trisaccharide in aqueous methanol and adding an equal volume of n-butanol along with a few drops of n-hexyl alcohol.]
3.3 Finally, D1 states (lines 19 to 22) that crystalline 2'-FL can also be obtained by using the crystalline products of the above two methods as seed crystals in the following method (denoted as "method 3" by the parties):
"Besitzt man Impfkristalle, so läßt sich die Fucosido-lactose (l g) durch Lösen in heißem 75-proz. Methanol (20 ccm) und allmähliche Zugabe von absol. Äthanol (60-80 ccm) leicht in schönen dreieckigen Plättchen (Abbild.) vom Schmp. 230-231° (Zers.) erhalten."
[translation by the board: If seed crystals are available, fucosido-lactose (1 g) can be easily obtained as nice triangular plates (see image) with a melting point of 230-231° (decomposition) by dissolving the crystals in hot 75% methanol (20 ccm) and gradually adding absolute ethanol (60-80 ccm).]
3.4 The above understanding of D1, i.e. that method 3 uses the crystals of method 1 or 2 as seeds, is consistent with the summary of D1 in paragraph [0003] of the patent and was also common ground between the parties at the oral proceedings before the board.
3.5 With respect to the crystalline 2'-FL obtained, D1 also states (lines 22 to 23) that it is non-hygroscopic.
This property of the crystalline 2'-FL is mentioned in D1 only after the description of method 3. Contrary to the appellant's view, however, this does not mean that this property characterises only the crystals resulting directly from method 3. In the present case, the crystals obtained by method 1 or 2 are used as seed crystals in method 3 (see above). This means that the crystals of method 3 are obtained from these seed crystals by further growth. Therefore, the board concurs with the respondent that it is not apparent why the crystals from method 3 should have different properties than those resulting from method 1 or 2. In summary, D1 can only be understood as meaning that the above property (non-hygroscopic) characterises the crystals of 2'-FL no matter which of the three disclosed methods is used.
4. D1 does not disclose XRPD reflections of the crystals from any method or the degree of their purity in w/w%. However, this does not mean that its disclosure cannot be novelty-destroying. To the extent that the teaching of D1 is sufficient for the skilled person, in light of their common general knowledge, to obtain the product of claim 1 (in other words: if D1 provides an enabling disclosure for the product of claim 1), D1 would be novelty-destroying to claim 1.
4.1 In this context, experiment 1 of D21 (page 2/20) is relevant. It reads as follows (text in square brackets added by the board):
"In order to repeat the Method 2 in Literature 1 [D1] the following experiment was carried out: 2.527 g of amorphous 2-FL [2'-FL] was dissolved in 100 ml of aqueous methanol (99 ml of anhydrous methanol + 1 ml of water), and the white solids were gradually dissolved under continuous stirring, the liquid became a white turbid liquid, and then 100 ml of n-butanol was added, and more white solids were precipitated. The solution was filtered to obtain a colorless transparent solution, and then 400 myl of n-hexanol was added to the filtrate, and shaken. The solution was transferred to a 250 ml conical flask, sealed with a parafilm perforated with several small holes, and stored at 4 °C for 2 weeks. The crystalline product was filtered, washed with methanol, and dried in vacuo to obtain 0.314 g of white crystals."
4.2 The board agrees with the respondent that experiment 1 of D21 is a proper repeat of method 2 of D1, i.e. a repeat which the skilled person would have carried out having their common general knowledge in mind. In particular, in both method 2 of D1 and experiment 1 of D21, amorphous 2'-FL is dissolved in a certain volume of aqueous methanol and the solution is diluted with the same volume of n-butanol. Afterwards, a few drops of n-hexanol are added and crystallisation is carried out at 4 °C.
4.3 As is evident from figure 4 and table 1 of D21 (pages 10/20 and 11/20), the crystals obtained from this repeat show the XRPD reflections recited in claim 1. This was also never disputed by the appellant.
4.4 Furthermore, as already set out above, D1 uses a 2'-FL as the starting material for its crystallisations which had previously been purified by repeated chromatography or via its tosylhydrazone. D1 (last sentence) also states that no other sugars could be detected in the mother liquors by paper chromatography after crystallisation.
Against this background, at least in the absence of any evidence to the contrary, which the appellant did not provide, the board is convinced that the crystalline 2'-FL obtained according to the disclosure of D1 and in particular method 2 does not contain any impurities.
5. The appellant submitted that D1 did not disclose method 2 in an enabling manner. In this respect, it referred to paragraph [0016] of the patent, stating that the inventors of the patent had never been able to reproduce the methods described in D1. Decisions T 325/16 and T 605/02 were relevant in this context. In particular, the appellant argued that several pieces of information were lacking in D1 so that especially method 2 might not be reproduced. Experiment 1 of D21 filled in these gaps in the disclosure of method 2 of D1. Contrary to D21, D1 did not specify the amounts of solvents or 2'-FL used nor did it exactly specify the number of drops of n-hexanol added or the storage time; moreover, D21 disclosed that the liquid became a white turbid liquid, an item of information that was not mentioned in D1. Therefore, several additional assumptions had been made in D21 that could not be derived from D1 nor from the skilled person's common general knowledge. The appellant also pointed to the fact that, according to D21, the crystalline product was obtained after only two weeks, whereas D1 disclosed the obtention of the crystals after "several weeks".
Furthermore, the appellant referred to D6 and submitted that it was evident from figure 1 of D6 (page 946) that several parameters such as e.g. choice of solvent, temperature, concentration, agitation and pH all had an effect on the type of polymorphic form that was obtained from a solution during crystallisation. Therefore, it was entirely conceivable that the polymorph as defined in claim 1 had been obtained in D21 by selecting appropriate operating conditions that were not mentioned in D1 but that only slight changes in the crystallisation conditions of experiment 1 of D21 would have resulted in a different polymorph.
Hence, the appellant argued that D21 was not a proper repeat of method 2 of D1 and therefore it could not be concluded that the 2'-FL polymorph of claim 1 was necessarily obtained when following the teaching of method 2 of D1.
6. The board is not convinced by these arguments for the following reasons.
6.1 It is clear that method 2 of D1 is not described to the very last detail. However, this does not allow the conclusion that the method is necessarily not disclosed in an enabling manner. In the present case, the board is convinced that the skilled person, having their common general knowledge in mind, would have known how to put method 2 of D1 into practice. The appellant's arguments are not suitable for casting doubt on this for the following reasons.
6.1.1 First of all, both the storage time at 4 °C (2 weeks) and the amount of n-hexanol (400 myL, i.e. about 8 drops) reported in D21 are fully in line with the disclosure of method 2 in D1 ("nach mehrwöchigem Aufbewahren" [translation by the board: after several weeks of storage], "sowie einiger Tropfen n-Hexylalkohol" [translation by the board: along with a few drops of n-hexyl alcohol]). In the absence of evidence to the contrary, which the appellant did not put forward, the board sees no reason to doubt that the skilled person would very well have chosen the storage time and number of drops of n-hexanol as stated in experiment 1 of D21.
6.1.2 Secondly, it is true that the amounts of solvents and 2'-FL used in experiment 1 of D21 are not disclosed in D1. However, the skilled person is well aware that any compound which is to be crystallised from a solvent or a solvent system has a certain solubility at a certain temperature in said solvent (system). Consequently, depending on the amount of compound to be crystallised, the skilled person would have used only those amounts of solvent(s) which result in a concentration of the compound to be crystallised which is above the solubility limit at the crystallisation temperature, because no crystallisation at all could otherwise have been achieved. In simpler terms, as argued by the respondent, the skilled person would have known how much solvent to use depending on the amount of 2'-FL to be crystallised. This is what has been done in experiment 1 of D21.
6.1.3 Thirdly, D21 also states that "the liquid became a white turbid liquid" upon dissolution of the amorphous 2'-FL in aqueous methanol. The appellant pointed to the fact that this observation is not described in method 2 of D1. This argument implies that according to the appellant, the solution in D1 must have been clear upon dissolution of the amorphous 2'-FL in aqueous methanol. However, in the absence of evidence to the contrary, which the appellant did not provide, the board sees no reason for this assumption. In fact, according to experiment 1 of D21 (loc. cit.), a clear solution is obtained by filtration. Thus, the board is satisfied that this is exactly what the skilled person would have done by following method 2 of D1 if they had wanted to prepare a solution from which a compound was to be crystallised.
6.2 Therefore, the appellant's arguments cannot change the conclusion above that method 2 of D1 is disclosed in an enabling manner and that experiment 1 of D21 is a proper repeat of method 2 of D1, i.e. a repeat which the skilled person would have carried out having their common general knowledge in mind.
6.3 In view of the above, the mere assertion in the patent that the appellant tried in vain to repeat the methods of D1 is not convincing. In fact, as submitted by the respondent, the appellant has not provided any details concerning the operating conditions that might have been used in these attempts to reproduce the methods of D1 and which would not have allowed crystallised 2'-FL to be obtained.
6.4 It may be that, as argued by the appellant with reference to D6, a slight modification of the crystallisation conditions of experiment 1 of D21 could have in principle resulted in a different polymorph. However, the allegation that such is exactly the case here has never been proven by the appellant and, thus, amounts to mere speculation.
6.5 The appellant also referred to decisions T 325/16 and T 605/02, arguing that a polymorph disclosed in the prior art in a non-enabling manner could not be novelty-destroying. The board had already pointed out the lack of relevance of these decisions in its communication pursuant to Article 15(1) RPBA 2020 (see point 3.4.8). This is because the underlying facts were different from the facts at hand. Moreover, as set out above, the board is convinced that the polymorph defined in claim 1 is disclosed in D1 in an enabling manner. At the oral proceedings before the board, the appellant no longer relied on these decisions in its argumentation. Hence, the board sees no reason to comment on them further.
7. Therefore, the board concludes that method 2 of D1 is disclosed in an enabling manner and that it results in the claimed 2'-FL polymorph without any impurities. Hence, the subject-matter of claim 1 of auxiliary request 2 is not novel over D1. Auxiliary request 2 is not allowable.
Main request and auxiliary request 1 - Claim 1 - Novelty (Article 54 EPC)
8. Claim 1 of auxiliary request 1 differs from claim 1 of auxiliary request 2 only in that it allows for a higher degree of impurity ("less than 10 w/w% of impurity" in claim 1 of auxiliary request 1 vs. "less than 5 w/w% of impurity" in claim 1 of auxiliary request 2). Claim 1 of the main request (claim 1 as granted) does not set any limit to the amount of impurity.
Therefore, the subject-matter of claim 1 of auxiliary request 2 is fully encompassed by the subject-matter of claim 1 of the main request and auxiliary request 1. The above reasoning of lack of novelty for the subject-matter of claim 1 of auxiliary request 2 therefore also applies to the subject-matter of claim 1 of the main request and auxiliary request 1.
It follows that the subject-matter of claim 1 of these requests is not novel over D1 and therefore the main request and auxiliary request 1 are not allowable.
