T 2049/19 (Ticagrelor formulation/ASTRAZENECA) 23-08-2021

1. The appeal is admissible. It meets the requirements of Articles 106 to 108 and Rule 99(2) EPC.

2. Absence of parties at the oral proceedings - Rule 115(2) EPC and Article 15(3) RPBA 2020

The oral proceedings before the board took place in the absence of appellant 2 and the party as of right (opponent 3). Both had been duly summoned. Therefore, in accordance with Rule 115(2) EPC and Article 15(3) RPBA 2020, the board decided to continue the proceedings in the absence of those parties.

The party as of right (opponent 3) had not presented any submissions in the course of the appeal proceedings. Appellant 2 was treated as relying on its written case. Hence, the board was in a position to announce a decision at the conclusion of the oral proceedings, in accordance with Article 15(6) RPBA 2020.

3. Inventive step (Article 56 EPC) - claim 1 as granted

3.1 The patent (paragraphs [0001], [0004] and [0006]) seeks to provide ticagrelor formulations for oral administration which release substantially all of the active ingredient.

3.2 The parties agreed that the tablets in Example 2 of document D3, especially Tablet II, constituted the closest prior art. The board sees no reason to differ.

D3 is cited in the patent (paragraph [0003]) and the corresponding passage of the application as filed. It concerns the preparation of crystalline and amorphous forms of ticagrelor, and pharmaceutical compositions comprising them (D3: abstract and page 1, lines 2-6; claims 1, 15, 18 and 29). In Example 2, D3 discloses Tablet II, which contains ticagrelor, lactose, croscarmellose sodium, maize starch, polyvinyl pyrrolidone and magnesium stearate.

3.3 Thus, the pharmaceutical compositions of claim 1 as granted and Tablet II of D3 differ in respect of their excipients.

3.4 It was a matter of dispute what technical effect(s) could be attributed to these differences. The respondent, relying on the comparative examples in post-filed test reports D13 and D25, asserted that the composition of claim 1 had improved stability and bioavailability. The appellants maintained that the improvements alleged by the respondent were not plausible from the application as filed. Therefore, post-filed evidence could not be taken into consideration. Moreover, they argued that the comparative tests described in D13 and D25 were not conclusive as they did not constitute a suitable comparison with the closest prior art.

3.5 Irrespective of whether the improvements alleged by the respondent were plausible on the basis of the application as filed and common general knowledge, the board agrees with the appellants that the comparative tests presented in D13 and D25 are not suitable for the purpose of showing any effect over the closest prior art, for the reasons set out below.

3.5.1 The following data (percentages calculated from the data on file) relate to the composition (tablet core) of Tablet II of D3 and Formulations 1 to 3 in D13. Tablet II is the formulation of the closest prior art. Formulation 3 of D13 is a composition according to claim 1. Formulations 1 and 2 are the comparative compositions used according to D13.

Function |Tablet II (D3) |Formulation 1 100 mg (D13) |Formulation 2 (D13) |Formulation 3 (D13) |

Pharm. active agent|Ticagrelor 16.7% |Ticagrelor (micronised) 28.6% |Ticagrelor (micronised) 30.0% |Ticagrelor 30.0% |

Filler |Lactose Ph. Eur.74.6% |Lactose monohydrate40.0% MCC 22.6% |Mannitol 42.3% DCPD 18.0% |Mannitol 42.0% DCPD 21.0% |

|Maize starch5.0%| | | |

Binder |PVP 0.8% |PVP 4.3% |PVP "K30"5.0% |HPC3.0% |

Disinte-grant |Croscarmellose sodium 2.0% |Croscarmellose sodium 4.0% |Croscarmellose sodium4.0% |Sodium starch glycolate 3.0% |

Lubricant |Magnesium stearate1.0%|Magnesium stearate0.5% |Magnesium stearate0.8%|Magnesium stearate1.0%|

In this table, MCC stands for microcrystalline cellulose, HPC for hydroxypropyl cellulose, DCPD for dibasic calcium phosphate dihydrate and PVP for polyvinyl pyrrolidone (povidone).

It was a subject of discussion whether the maize starch used in Tablet II should be included in the category of filler rather than binder. As pointed out by the respondent, Example 2 of D3 does not indicate that the maize starch in Tablet II was pre-gelatinised ("starch paste"), a required step for it to act as a binder (reply to the statements setting out the grounds of appeal: page 10, paragraph 1, and D11: page 861, left-hand column, paragraph 3).

3.5.2 The respondent relied on the following data.

(a) D13 describes in vitro dissolution tests of Formulations 1 to 3 in a dissolution medium, presumed to correlate with the bioavailability of ticagrelor (see D13, page 4, paragraph 1). According to the results shown in the figure below, Formulations 2 and 3 release ticagrelor to a similar extent, which is greater than that achieved with Formulation 1.

FORMULA/TABLE/GRAPHIC

(b) The respondent stated furthermore that the thus predicted higher bioavailability of Formulations 2 and 3 was confirmed in vivo in a bioequivalence study (D13: paragraph bridging pages 4 and 5).

Formulation 2 achieved higher values of Cmax and AUC than Formulation 1, as shown in the following figure (included on page 15 of the respondent's reply to the statements setting out the grounds of appeal). The curve designated Phase IIb corresponds to Formulation 2, and that designated Phase I/IIa to Formulation 1.

FORMULA/TABLE/GRAPHIC

(c) Lastly, it is mentioned in D13 (page 5, last full paragraph) that accelerated stability studies showed that Formulation 3 had a superior stability profile to Formulation 2.

3.5.3 The respondent argued that although Formulation 1 differed in several technical features from Tablet II of D3, it was suitable as a comparator for showing an improvement over the closest prior art for the following reasons.

(i) Ticagrelor in Formulation 1 was micronised, so it should dissolve faster than in Tablet II, where it was not micronised (D11, right-hand column, paragraph 3).

(ii) Formulation 1 contained double the concentration of disintegrant of Tablet II, so it should also disintegrate faster.

(iii) Other differences between Formulation 1 and Tablet II, especially those relating to the filler, were immaterial.

Regarding points (i) and (ii), as the disintegration and dissolution rates were directly linked to bioavailability, Formulation 1 could be expected to provide higher ticagrelor bioavailability than Tablet II. Thus, as the bioavailability of the formulation according to claim 1 (Formulation 3) was superior to that of Formulation 1, it was also superior to Tablet II.

Furthermore, the sequence Tablet II > Formulation 1 > Formulation 2 > Formulation 3 depicted the timeline in the development of ticagrelor formulations by the respondent, which led to the commercialisation of the product Brilique**(®). Formulation 1 had resulted from the optimisation of Tablet II, to be used in Phase I/IIa tests. Formulation 2 was used for Phase IIb clinical studies, and Formulation 3 was the optimised formulation for Phase III studies and commercialisation. Hence, it was evident that Formulation 1 must be superior to Tablet II.

Regarding point (iii), the appellants had not shown that differences between Tablet II and Formulation 1 other than (i) and (ii) were relevant.

3.5.4 These arguments cannot succeed for the following reasons.

While this was not shown for the actual formulations in question, the assumption that micronisation and an increase in the amount of disintegrant may result in higher dissolution and disintegration rates appears plausible (points 3.5.3(i)-(ii) above).

However, it cannot be assumed without any evidence that the other remaining differences between Formulation 1 and Tablet II (point 3.5.3(iii) above) are immaterial.

The respondent argued that replacing the filler material of Tablet II by the filler of Formulation 1, i.e. an equivalent excipient having the same function in the formulation, would not make a difference.

Yet when comparing Formulations 1 and 2, which both contain micronised ticagrelor and essentially the same combination of binder and disintegrant, the respondent came to the conclusion (D13, page 4, last full paragraph, and reply to the statements of grounds of appeal, page 15, last paragraph) that replacement of the filler lactose/microcrystalline cellulose by the filler mannitol/DCPD resulted in an unexpectedly superior formulation. In other words, the respondent attributed the enhanced dissolution rate and bioavailability of Formulation 2 to the filler composition and considered that this effect could not have been predicted.

Given the unpredictable effect of a modification of the filler on dissolution and bioavailability, it is uncertain what would have been the impact of reducing the total filler content in Tablet II from 79.6% to 62.6%, removing maize starch, and replacing a considerable proportion of the lactose by microcrystalline cellulose.

In addition, the effect of increasing the povidone content to more than five times the amount, from 0.8% to 4.3%, is also unknown.

Therefore, the board cannot conclude that Formulation 1 necessarily provides higher bioavailability than Tablet II. Consequently, the conclusions drawn from comparative tests relating to Formulation 1 cannot be extended to Tablet II; the tests reported in D13 do not show that the claimed composition exhibits higher bioavailability than the composition of the closest prior art.

3.5.5 The respondent's argument that Formulation 1 necessarily had superior bioavailability to Tablet II because it represented a more advanced stage in the development of ticagrelor formulations is not convincing. As argued by appellant 1 at the oral proceedings before the board, the use of Formulation 1 in Phase I/IIa tests rather than Tablet II was not necessarily linked to improved bioavailability; it could have been due to other factors, for instance the cost of the preparation process. In the absence of supporting evidence, attributing the selection of Formulation 1 to higher bioavailability than Tablet II is a speculative exercise. Therefore, the related argument cannot be taken into consideration.

3.5.6 Regarding the aspect of stability, the respondent observed that replacing the binder and disintegrant of Formulation 2 by those in Formulation 3 resulted in an improvement. Although the nature of the binder and disintegrant in Tablet II is the same as in Formulation 2, their proportions are different. But more importantly, a comparison between Formulations 2 and 3 does not take account of the impact of the filler on tablet stability. Hence, this comparison is not suitable for establishing any improvement over Tablet II either.

3.5.7 The comparative tests in D25 show that replacing polyvinyl pyrrolidone by hydroxypropyl cellulose in a specific formulation results in improved disintegration times. As in D13, this comparative test was not carried out in relation to Tablet II of D3, and fails to establish the effect of replacing components other than the binder, e.g. the filler, and of their different amounts. Therefore, the tests in D25 are also inconclusive.

3.6 It follows from the above that the respondent failed to prove the alleged technical effects of the subject-matter of claim 1 over the composition of the closest prior art (Tablet II of Example 2 of D3).

Hence, the objective technical problem to be solved has to be formulated as the provision of an alternative formulation of ticagrelor. In this context, an alternative formulation means a further formulation suitable for the oral administration of ticagrelor.

3.7 The board is satisfied that the pharmaceutical composition of claim 1 as granted solves the objective technical problem. This was never contested by the appellants.

3.8 It was common ground that the fillers, binders and disintegrants in Tablet II of D3 and in claim 1 as granted were generally known on the priority date of the patent. This was apparent from the handbook excerpts collected in documents D9 and D11, which in the board's view represent common general knowledge.

Lactose, starch, mannitol and dibasic calcium phosphate dihydrate were widely used as tablet fillers (D9: page 385, section 7, page 389, section 7, page 449, section 7, and page 96, section 7; D11: page 860, left-hand column, last paragraph). Polyvinyl pyrrolidone and hydroxypropyl cellulose were widely used as tablet binders (D9: page 611, section 7, and page 336, section 7; D11: page 861, left-hand column, paragraphs 5 and 6). Moreover, croscarmellose sodium and sodium starch glycolate were known as tablet "super disintegrants" (D11: page 882, left-hand column, paragraph 5; D9: page 211, section 7, and page 701, section 7).

Taking into account common general knowledge, it would thus have been obvious, to the skilled person searching for another formulation suitable for the oral administration of ticagrelor, to replace some or all of the excipients in Tablet II of D3 by excipients known to fulfil the same function. Thus, replacement of the filler lactose/maize starch by mannitol/DCPD, the binder polyvinyl pyrrolidone by hydroxypropyl cellulose and the disintegrant croscarmellose sodium by sodium starch glycolate did not require inventive efforts.

3.9 Consequently, the subject-matter of claim 1 as granted does not involve an inventive step and the ground for opposition of Article 100(a) EPC in combination with Article 56 EPC prejudices the maintenance of the patent as granted.

4. Inventive step (Article 56 EPC) - auxiliary requests

In the written proceedings, the inventive-step arguments put forward by the respondent in relation to the auxiliary requests were based on an alleged improvement of the claimed compositions over the closest prior art (see the annex filed with the reply to the statements of grounds of appeal, section 4).

At the oral proceedings, the board announced its conclusion that the comparative examples in D13 and D25 were not conclusive and that the subject-matter of claim 1 as granted did not involve an inventive step. It then expressed its preliminary opinion that these conclusions also applied to auxiliary requests 1 to 8. The respondent nevertheless chose not to present any arguments as to why the auxiliary requests would overcome the lack of inventive step of the main request.

Compared to claim 1 as granted, the corresponding independent claims of the auxiliary requests (see point VI above) indicate as further restrictions that the composition is for oral administration (auxiliary requests 1, 4, 5 and 7) and specify concentration ranges for some or all of the ingredients in the composition (auxiliary requests 2 to 8).

Tablet II in Example 2 of D3 is a tablet for oral administration (D3, page 12, lines 16-21). Hence, the limitation to oral administration in claim 1 of auxiliary requests 1, 4, 5 and 7 does not constitute an additional difference to the closest prior art and cannot contribute to inventive step.

Regarding the concentration ranges specified in the respective claim 1 of auxiliary requests 2 to 8, they appear to fall within the customary concentration ranges known for each of the ingredients when fulfilling their respective function in tablet formulations (see for example D9), and have not been shown to produce any unexpected effect. The concentration range indicated for ticagrelor is also a customary modification with no associated unexpected effect.

Hence, the amendments made to claim 1 in the respective auxiliary requests do not overcome the lack of inventive step of the composition of claim 1 as granted. Consequently, none of auxiliary requests 1 to 8 meets the requirements of Article 56 EPC.