T 0018/10 (Caries Lectin detection/UNIVERSITY SOUTHERN CALIFORNIA) 09-09-2013

Reasons for the Decision

Admissibility of the Main Request

1. Except for a clerical correction in claim 5, namely to define lectin UEA as UEA I, the Main Request filed with the appellant's statement of Grounds of Appeal on 16 November 2009 is identical to the request filed on 20 May 2009, the sole claim request considered by the examining division in the decision under appeal. The Main Request is thus admitted into the appeal proceedings.

Procedural issues

2. According to decision G 10/93 (OJ EPO 1995, page 172, Headnote) "In an appeal from a decision of an examining division in which a European patent application was refused the board of appeal has the power to examine whether the application or the invention to which it relates meets the requirements of the EPC. The same is true for requirements which the examining division did not take into consideration in the examination proceedings or which it regarded as having been met. If there is reason to believe that such a requirement has not been met, the board shall include this ground in the proceedings".

In the present case, the board considers that such additional grounds exist and therefore, in its communication pursuant to Article 15(1) RPBA, informed the appellant of these additional grounds (cf. Section VI supra). In view of the fact that the appellant has not provided any substantive reply to these grounds (cf. Section VII supra), the present decision is essentially based on the board's objections raised in this communication.

Main Request

Article 84 EPC; Articles 84 and 83 EPC

3. According to the established case law of the Boards of Appeal, there is no reason to interpret an excessively broad claim more narrowly, if it is a question not of understanding concepts that require explanation but rather a question of examining an excessively broad request in relation to the state of the art (cf. "Case Law of the Boards of Appeal of the EPO", 6th edition 2010, I.C.2.9, page 105). The claims must also reflect the actual contribution to the art in such a way that the skilled person is able to perform the invention over the entire range claimed (cf. "Case Law", supra, II.B.4.1, page 275). In view of this case law, the board considers that some claims do not fulfil the requirements of Article 84 EPC, in particular as regards clarity and support by the description.

4. The method of claim 1 for predicting the risk of dental caries in a subject comprises three steps, namely: 1) contacting an aliquot of an unfractionated saliva sample - containing more than one lectin-binding component - with more than one lectin, under conditions that allow the binding of the lectins to the respective lectin-binding component, 2) detecting the amounts of lectins bound, and 3) comparing these amounts to the amounts known to bind a saliva sample from a control subject, wherein the amounts of bound lectins are indicative of the risk of dental caries in the subject (cf. Section II supra).

4.1 The method of claim 1 embraces embodiments which rely on the combination of two lectins only, without a limitation to any particular type of lectin. In view of the actual disclosure of the application, the board considers that there is no technical support for these embodiments.

4.1.1 In Example I of the application, the most predictive lectins related to the MUC7 and MUC5B mucin concentrations (known in the art to have the best association with the forecast of decay and filled permanent teeth, DFT) are identified (cf. page 49, lines 3-6). MAL I (Maackia Amurensis Lectin I; for alpha-2,3 linked sialic acid) has the highest individual correlation with DFT (cf. page 50, Table 1 and page 52, lines 15-17) and, in view of the results obtained, ACL (Amaranthus Caudatus Lectin; for different configurations of sialylated T-antigen) was dropped from the lectin panel (cf. page 50, lines 21-22). The prominence of sialic acid is further supported by the results obtained with SNA (Sanbucus Nigra Lectin; primarily for alpha-2,6 linked sialic acid). JAC (Jacalin, Jackfruit seed lectin; for a different configuration of sialylated T-antigen) identifies a form of the T-antigen that may be a negative factor (cf. page 50, lines 19-21). AAL and UEA I (Aleuria Aurantia Lectin and Ulex Europaeus Lectin I; for different Lewis antigens, especially the variety with alpha-1,2 linked fucose) are "somehow ... interwoven in the relationship between mucin and DFT" (cf. page 50, lines 17-19). The outcome of the overall relationship of DFT to a panel of lectins containing MAL, JAC and MAA (Maackia Amurensis), after standardizing the contribution of each independent variable, is acknowledged to be complex (cf. page 53, lines 21-27).

4.1.2 In line therewith, the test composition used in Example 2 is based on "the lectins that have been shown to be contributors to the predictive regression equations" in Example 1, with reference to three lectins (cf. page 59, lines 4-19). In Example 3, a combination of the three lectins MAL I, JAC and SNA (disclosed in Table 1 as contributors) is used for assigning tested individuals to four different groups: high, medium, low and zero DFT (cf. page 62, lines 23-26). When combined with other factors (age, ethnicity), it is found that "MAL I accounts for more than 50% of the regression equation" (page 63, lines 7-8). In assays with further lectins, only slightly better results are obtained (R**(2) improves from 0.926 to 0.990 in the best case, with possible alternatives 0.957 and 0.966, all of them > 0.90). In all these assays, MAL I and JAC (the two lectins identified in Table 1 as having the most relevant results in inverse correlation to MUC5B and MUC7) are used (cf. page 64, lines 15-24).

4.1.3 The results obtained with individual lectins and combinations thereof (cf. page 50-52, Tables 1-3) support the following conclusions: 1) each lectin has a particular DFT correlation, 2) not all lectins have a significant DFT correlation, 3) MAL I provides the strongest link to DFT forecast (cf. page 54, lines 7-8), and 4) certain combinations of lectins do not improve the strength of the relationship (of MAL I) to DFT, only a few combinations improve the DFT correlation (cf. page 52, line 17 to page 53, line 27).

4.2 As regards the Appendix filed by applicant/appellant on 20 October 2008, the following points are of relevance:

4.2.1 The data based on multiple linear regression is obtained with a mixture of at least 12 lectins, except for one example with 9 lectins, in which the number of lectins was lowered by using "a proprietary data management system". The formulation date of these mixtures is sometime in 2005, two years after the priority date of the application.

4.2.2 The lectins used in the mixtures analyzed by multiple linear regression have different specificities, such as alpha-2,3 (MAA), alpha-2,6 (SNA), alpha-1,2 (AAL), and some of them include the lectin JAC, shown in the application to have an inverse relationship or correlation to other lectins. The presence of ACL, VVA and/or GNL in some of these mixtures does not provide much information, since they all consist of a large number of lectins and thus, their contribution may not be relevant or substantial for the results obtained.

4.2.3 The data based on neural nets have also been obtained in the years 2005-2008 and refer to a "combination of proprietary data management system and neural net mathematics" which are not disclosed in the application. Except for the mixtures indicated in point 10 of the Appendix, all other mixtures contain at least 5 lectins and have, prima facie, properties (specificity) similar to those used for multiple linear regression. Thus, the presence of VVA in some of them may not be relevant for the results obtained.

4.2.4 Point 10 of the Appendix is the sole source of information on file concerning combinations of two lectins. However, the correlation given for these combinations with the risk of caries in young adults is very low (p=<0.05; see for comparison Table 3, page 52 of the application) and no information is provided for each of them separately. Two of these mixtures contain ACL, which was dropped from the lectin panels used in the Examples of the application (supra).

4.2.5 The relevance of the information provided in this Appendix is questionable since it is not a complete disclosure. There is no information regarding the methods and parameters used which may be of relevance for achieving the information disclosed, such as, inter alia, the proprietary data management and the neural net.

4.3 The results shown in the Appendix are in line with those obtained in the Examples of the application which, as stated in point 4.1.3 supra, allow the following conclusions: 1) the contribution of MAL I and JAC to the DFT correlation is substantial and highly relevant, 2) the presence of additional lectins with different specificities may optimize this correlation; however 3) the contribution of other lectins may be irrelevant. In view thereof, the board considers that a generalization to use any possible combination of whatever lectins (more than one, at least two) in a method for predicting the risk of dental caries is not supported by the disclosure of the application.

4.4 It is arguable whether, in the method of claim 1, the lectins have to be used simultaneously (as a mixture) (as argued by the appellant) or whether they may be used spatially and/or temporally separated (as argued by the examining division). This interpretation is of relevance for defining the characteristics and properties of the assay device of claims 18-21 (cf. Section II supra). The references found in the application, such as on page 23, line 31 to page 24, line 1 and on page 28, lines 21-23 and the fact that, for a mixture of three lectins (MAL I, JAC and SNA), three different Western blots were used (cf. paragraph bridging pages 63 and 64), are considered by the board to support the interpretation of the examining division.

4.5 The definition of the term "control subject" given in the description introduces a certain degree of ambiguity in claim 1. On the one hand, a control subject is merely a subject without the disease (dental caries), such as indicated on page 19, lines 10-12, page 20, lines 14-15 and page 26, lines 19-21, on the other hand, a control subject is also defined as a subject with a known disease level, in particular a subject classified - by an alternative method (number of cavities) - as belonging to a very low risk group, such as indicated in the paragraph bridging pages 23-24 and page 44, lines 22-29. It is noted that in Examples 1 and 2, control subjects are identified by their levels of MUC7 and MUC5B mucin concentrations.

4.6 Likewise, in view of the different definitions given to the term "control subject", the reference in claim 1 to the amounts of bound lectins as being indicative of the risk of dental caries is ambiguous. The more so in the light of claim 2, which implies that the amounts indicated in claim 1 may be neither above nor below those of a control subject.

4.7 In view of the sheer number of possible lectin combinations and the scarce information provided by the application, the board considers that claim 1 is drafted in terms of the result desired to be achieved (an objection raised by the examining division against the subject-matter of claim 27, see page 4, point 5.3 of the decision under appeal) and that the application does not provide sufficient technical support to perform the invention over the entire scope of claim 1 (cf. point 4.3 supra).

5. The subject-matter of claim 5 includes the lectin ACL which, according to Example 1 of the application, is not significant and was thus dropped from all lectin panels used (cf. page 50, lines 21-22 and Table 1). Apart from the Appendix filed on 20 October 2008, whose deficiencies are discussed in point 4.2 supra, there is no support on file for combinations of two lectins, one of them being ACL.

5.1 It is also noted that none of the lectins referred to in claim 5 or in any of the other dependent claims is designated with a complete name but only in abbreviated form. It is questionable whether each and every one of these abbreviations may be clearly understood by a skilled person as being standard and well-known in the field, i.e. without introducing any ambiguity into the claims. Likewise, a similar objection applies to abbreviations concerning dental diseases, such as those referred to in claim 27 (cf. Section II supra).

6. The use of claim 14 relies on a therapeutic reagent which is defined as being "effective for dental caries" (cf. Section II supra). In view thereof, the following questions arise:

6.1 It is questionable whether all therapeutic reagents effective for the treatment or reduction of dental caries are also efficient in the prevention of dental caries (see page 41, "Anti-Caries Reagents").

6.2 In view of the references in the application to the "identification of population risk" when using other methods, such as on page 23, lines 14-18 and page 44, lines 22-29, the question arises whether the population group identified by the method of claim 1 is actually a new subpopulation and whether this subpopulation essentially overlaps, completely or partially, with those identified with these other methods known in the art. In other words, does the scope of claim 14 embrace a known use (preventing/reducing the risk of dental caries) for known products (anti-caries reagents) in a subpopulation which is also known in the art (medium and high risk) but which is now only identified by a different method?

6.3 Since the appellant has failed to provide a substantive reply to these questions (cf. Section VII supra), the board is not in a position to acknowledge claim 14 to fulfil the requirements of Article 84 EPC alone and/or in combination with Article 83 EPC.

7. The subject-matter of claim 15 does not relate to the prevention or reduction of dental caries risk. The purpose-limited kit is explicitly stated to be "for detecting dental caries" (cf. Section II supra). In view thereof, the following issues are of relevance:

7.1 The claimed kit is required to comprise "means for measuring the amounts of more than one lectin-binding components". However, the means are not - necessarily and exclusively - those referred to in the method of claim 1, i.e. "more than one lectin under conditions ...". They may well be completely different and not rely on binding but on other techniques which are not suggested, let alone disclosed, in the application. Moreover, it is not evident whether means useful for predicting the risk of dental caries must, always and necessarily, be identical to those useful for detecting dental caries.

7.2 Indeed, the term "means" itself is ambiguous, since it may refer to a technical device for carrying out a measure (cf. page 20, lines 19-22) or else to the products required for identifying and measuring the lectin-binding components. The nature and composition of the "oral fluid standard" is also ambiguous in view of the definition given in the application which, as stated on page 25, lines 22-31 may include a spiked surrogate and may be dependent on the context (cf. page 39).

8. Likewise, the subject-matter of claim 18 and dependent claims 19-21 neither relates to the prediction of dental caries nor to the detection of dental caries, but only to a mere detection of the presence of (whatever) "more than one lectin-binding component in a saliva sample" (cf. Section II supra). There is no requirement that the (more than one) lectins bound to a first matrix material in the receiving zone have to be of relevance for the prediction and/or detection of dental caries. In the absence of such a requirement, the nature and composition of the (at least one) "control saliva sample of a known concentration" in the control zone is also considered to be completely ambiguous.

9. The subject-matter of claim 27 refers to a first and a second set of (two or more) lectins that are positively and negatively correlated with one or more of several dental prognosis, respectively. In view of the comments made above and the scarce information provided by the application (Tables 1-3), the board, as stated in point 4.7 supra, considers that the objection raised by the examining division on page 4, point 5.3 of the decision under appeal is justified (cf. Section III supra).

10. In view of all the foregoing observations and objections, the board considers the Main Request not to fulfil the requirements of Article 84 EPC alone and/or in combination with Article 83 EPC.

Auxiliary Requests I and II

11. Auxiliary Request I filed with the appellant's statement of Grounds of Appeal differs from the Main Request by the deletion of claim 27. Auxiliary Request II also filed with the appellant's statement of Grounds of Appeal differs from the Main Request in that the subject-matter of claim 5 of the Main Request has been incorporated in claim 1. In view of the findings below, it is not necessary to assess the admissibility of these requests in the appeal proceedings.

12. Many of the objections raised above for the Main Request under Article 84 EPC alone and/or in combination with Article 83 EPC apply also to Auxiliary Requests I and II. Therefore, Auxiliary Requests I and II do not fulfil prima facie the requirements of the EPC.