[Note from Author - In my original post, I characterized the patentholder's victories as modest. It has since been pointed out to me that frameshift mutations account for a large percentage of the relevant mutations inBRCA1, so the most recent decision allowing claims directed to methods of detecting frameshift mutations in EP699754 could constitute more than a modest victory for the patentholder. Since the decision has not been published, and I am only going off of the EPO press release, it is impossible to know the actual scope of the remaining claims. We will have to await the decision before making any sort of real conclusion as to the import of the remaining claims.]
On November 19, 2008, a Technical Board of Appeal (TBA) of the European Patent Office (EPO) decided that EP 699754, a European patent relating to the BRCA1 gene and methods for diagnosing a predisposition for breast and ovarian cancer by identifying mutations in the gene, is to be maintained in an amended form. The patent is assigned to the University of Utah, but it is widely reported that Myriad Genetics is a real party in interest – Myriad was initially an assignee of the patent, but later the designation of Myriad as a proprietor was removed from the patent, perhaps owing to the widespread negative feelings towards Myriad in Europe which resulted from a perception that the company was engaging in overly aggressive enforcement tactics.
The patent was opposed by a number of groups from the public and private sectors of various European countries, including research institutes in various national centers for human genetics, and the initial opposition proceedings had resulted in a full revocation of the patent in 2004. On appeal, the TBA set-aside the initial opposition decision and reinstated the patent, albeit in an amended form.
The TBA decision and the text of the amended claims do not appear to have been posted on the EPO website yet, according to an EPO press release the claims have been narrowed to encompass only diagnostic methods for the detection of a predisposition for breast and ovarian cancer caused by a specific group of mutations of the gene, so-called frame-shift mutations. The patent in its current form does not contain claims directed to the BRCA1 gene itself or to mutated forms thereof. According to the press release, the patent cannot be further contested at the European level, indicating that the TBA rejected any requests by the University for referral to the Expanded Board of Appeals (EBA). The patent will still be susceptible to challenge at a national level.
This is the third case in which the University of Utah has succeeded in maintaining European BRCA1 patents, but in each case the resulting claims are significantly narrower than both the originally issued European claims and Myriad’s corresponding BRCA1 patents in the US.
A week earlier, on November 13, 2008, the EPO issued another press release reporting that another University of Utah BRCA1 patent, EP 705903, is likewise to be maintained in an amended form resulting in much narrower claims than the original patent. In this case, the TBA affirmed the result of the initial opposition proceedings, which resulted in the revocation of all diagnostic method claims, leaving only claims relating to a “gene probe of a defined composition for the detection of one specific BRCA1 mutation,” reportedly relating to a mutation specific to Ashkenazi Jewish populations. The actual TBA decision has not been posted yet, but should be in due course.
The first TBA decision relating to one of the universities BRCA1 patents (T 1213/05) was decided September 27, 2007, and resulted in the maintenance of EP 705902, albeit with substantially narrowed claims. Initially, the patent included a claim directed to the full-length BRCA1 gene, and another claim broadly encompassing hybridization probes capable of recognizing BRCA1 mutations. The claims ultimately allowed by the TBA are limited to a single probe capable of recognizing one exon of BRCA1, and several other probes capable of recognizing some BRCA1 intron-exon junctions. It is unclear how relevant these probe claims are to BRCA1 testing in its current state, but I suspect that the claims could be readily avoided by the use of diagnostic testing procedures that do not rely on the specific probes. I think it will be difficult to argue for an expansive interpretation of the claims, particularly in view of emphatic statements in the TBA decision to the effect that the claims are limited to probes (emphasis in the TBA decision), and the invalidation of the claims reciting the full-length gene and the broad genus of probes capable of detecting mutation.
The TBA decision regarding EP 705902 is available through the EPO website, and includes a number of interesting features. For one thing, sequencing errors in the initial US patents applications prevented the university from relying on these applications as priority documents. Unable to benefit from the earlier priority dates, the TBA found that the broader claims were anticipated by the inventors’ own publication of the correct sequence in Science prior to filing a US patent application including the correct sequences. In the US, with its one-year grace period, publication less than one year before the application filing date will not stand as a bar to patentability. But in Europe, which lacks any grace period, publication prior to the filing of a patent application results in a lack of novelty. This decision highlights the vulnerability of any DNA patent, particularly in Europe, which is based upon a patent application containing a sequencing error. This could be a significant issue, given the relatively high error rate in DNA sequencing. Of course, this error rate depends upon the methodology used for sequencing, when the sequencing was performed, the expertise of the party performing the sequencing, etc.
The opponents of EP 705902 argued that the patent is invalid for violating the “ordre public,” or morality, as referred to in Article 53(a) ofteh European Patent Convention (EPC). In particular, they argued that the patents were immoral based on the university’s failure to provide proof that the patients whose genetic mutations had been instrumental in identifying the gene had provided informed consent to use the information to secure a patent. The TBA rejected this argument, finding that there is no obligation for a patent applicant to provide evidence of informed consent.
The opponents also argued that the patent is immoral because of the social and economic consequences, i.e., the patent would only result in increased costs for patients, would influence the way in which diagnosis and research will be organized in Europe, and would clearly be to the detriment of patients and doctors. The TBA responded that 53(a) pertains to cases where “exploitation of the invention” would be immoral, not “explication of the patent.” Since the objections were raised against the patent rather than the underlying invention, the board determined that the EPO was not the proper venue for considering the socio-economic effects of a patent directed to a legitimate invention.
On t he other hand, the TBA also rejected challenges to the narrow claims which did survive the initial opposition proceedings. For example, the opponents of the patent argued that the probes recited in the university’s surviving claims were invalid for lack of novelty, based on a prior art reference disclosing a yeast artificial chromosome (YAC) allegedly including exon 11 and other exons of BRCA1. Claim 1 in the patent surviving opposition recites a probe comprising a sequence fragment derived from exon 11. However, the board rejected this argument, finding that the opponents had failed to establish with a sufficient degree of certainty that the YAC clone actually contained any of the sequences specified in claim 1.
The opponents also argued that the claims allowed after opposition lacked an “inventive step,” i.e., that the claims were obvious in view of a prior art reference which identified the location of a breast cancer gene in an approximately 1.5 Megabase (Mb) region of a human chromosome. This turned out to be the location of BRCA1. The opponents argued that once the location of the gene had been mapped with this degree of specificity, it only involved routine experimentation to clone and sequence the gene. They alleged that the university had won the race simply due to working harder and putting more money into the project. However, the TBA rejected this argument, finding that even after the general location of the gene had been mapped, finding the mutation “would not only involve a substantial amount of work, but would also require a ‘lucky strike’, which could in no way be predicted” based on the prior art.
These three European BRCA1 decisions are significant for at least two reasons. First, although the University of Utah (and Myriad) can claim partial victories, it is unclear how relevant the resulting patents are, and to what extent they impose meaningful restrictions on competing providers of BRCA1 testing in Europe. And secondly, the decisions provide some insight into legal arguments that could be made in the US if the BRCA1 patents were ever challenged in court. As I reported in a recent Science article (Christopher M. Holman, “Trend in Human Gene Patent Litigation,” Science, 322:198-99 (2008)), Myriad has only asserted its BRCA1 patents in two lawsuits, both filed more than 10 years ago. Both cases were quickly settled and dismissed, prior to any substantive decisions pertaining to the scope or validity of the claims. The EPO decisions at least point out some potential vulnerabilities of the much broader US BRCA1 patents, and perhaps gene patents in general.
For more information, follow this link for an EPO press release, which provides information on these decisions and links to other relevant press releases.
Saturday, November 22, 2008
Tuesday, November 4, 2008
Federal Circuit Affirms Rejection of Claim Reciting Broad Genus of Monoclonal Antibodies For Failure to Satisfy Written Description Requirement
In In re Alonso, decided Oct. 30, 2008, the Federal Circuit affirmed a Board of Patent Appeals and Interference (“Board”) decision sustaining a written description rejection of a claim reciting a method for treating a neurofibrosarcoma (a type of cancerous tumor) using a human hybridoma-derived monoclonal antibody idiotypic for the patient’s tumor, i.e., a human monoclonal antibody capable of recognizing a particular epitope on that patient’s tumor. The Court held that the applicant’s disclosure of a single monoclonal antibody capable of recognizing one particular patient’s neurofibrosarcoma failed to adequately describe the recited genus purporting to encompass all human hybridoma-derived monoclonal antibodies capable of recognizing any patient’s neurofibrosarcoma.
The PTO routinely allows claims broadly reciting a genus comprising any isolated antibody capable of recognizing a specified antigen, without limiting the applicant to any specific antibody, or any specific epitope on the antigen (a single antigen typically has multiple epitopes, the specific and discrete region of the antigen bound by a given antibody). The PTO's revised written description training materials, published earlier this year, specifically states that the disclosure of a reasonably well characterized antigen satisfies the written description requirement with respect to a claim encompassing any isolated antibody capable of binding the antigen (see earlier post discussing the training materials). This is consistent with the Federal Circuit's decision in Noelle v. Lederman, which found that although the disclosure of a mouse antigen was insufficient satisfy the written description requirement with respect to a claim reciting a monoclonal antibody capable of recognizing the human counterpart of the mouse antigen, disclosure of the mouse antigen would provide adequate written description support for a claim broadly reciting any antibody capable of recognizing the mouse antigen. It is worth pointing out that under this approach, not only is it unnecessary for the patent applicant to disclose the structure of any monoclonal antibody falling within the scope of the claim, it is also unnecessary to disclose any information regarding the structure of the antigen. This is inconsistent with other Federal Circuit written description cases relating to gene sequences which have required some disclosure of chemical structure, i.e., Fiers v. Revel, UC v. Lilly, In re Wallach, and most recently Carnegie-Mellon v. Roche (discussed in an earlier blog post).
Alonso’s claim is distinguishable, however, because it is not directed to a single antigen. The PTO and Court found that this sort of tumor is characterized by “considerable antigenic heterogeneity,” both between patients and between metastatic sites within a single patient. It was this unpredictability with respect to the structure of neurofibrosarcoma antigens, and consequently the genus of antibodies recited in the claim, which was the basis for the written description rejection.
In Alonso, the PTO is employing the written description requirement as a doctrinal tool for limiting claim scope. Traditionally, the enablement requirement has played the primary role in policing claim scope. For example, in Amgen v. Chugai the enablement requirement was invoked to invalidate a claim broadly reciting genetic sequences encoding structurally undefined functional analogs of human erythropoietin, with the court essentially finding that the claimed genus exceeded the scope justified by Amgen's disclosure of a single gene encoding wild type erythropoietin. But in UC v. Lilly the Federal Circuit established a new role for the written description requirement, which I refer to as the Lilly written description requirement (LWD), and which many commentators have characterized as a super-enablement requirement. Indeed, the examiner rejected Alonso’s claim for failure to comply with both the enablement and written description requirements, treating the two doctrines as effectively redundant. The Board, however, only affirmed the written description rejection, reversing on the issue of enablement.
The examiners tandem enablement-LWD rejection was not surprising. For years, examiners have asserted violations of both the enablement and LWD requirements in rejecting what they perceive to be overly broad genus claims. But until recently, it was almost unheard of for the Board to explicitly overrule the enablement rejection while affirming on LWD. In fact, when I conducted a comprehensive study of LWD cases a couple of years ago I found only one case where the Board (or any court for that matter) reversed an enablement rejection but affirmed on LWD. That decision, Capon v. Esshar, was soundly reversed by the Federal Circuit on appeal. In its decision, the Federal Circuit stated that analysis for compliance with the enablement and LWD requirements are essentially coextensive, and suggested that it would rarely if ever be proper to find a claim enabled but nevertheless in violation of LWD. Thus, in this respect Alonso breaks new ground.
The Federal Circuit has repeatedly stated that an applicant can satisfy LWD with respect to a genus claim by disclosing "a representative number of species in that genus." In Alonso the court agreed with the board that, in view of the heterogeneity and unpredictability of the antigen, disclosure of a single species did not constitute a “representative number." However, the decision provides no guidance with respect to how many species would represent a representative number. In fact, although the "representative number" standard was set forth in UC v. Lilly in 1997, a cited repeatedly by the courts and PTO, to my knowledge neither the PTO nor the courts have yet to provide any useful guidance with respect to how to determine just how many species is necessary to constitute a "representative number.” In fact, I would argue that the test for compliance of a genus claim with LWD cannot be meaningfully distinguished from the test for enablement, as suggested by various panels of the Federal Circuit, e.g., Capon v. Esshar and Lizardtech.
Alonso exemplifies what I see as a recent emphasis in the patent office on carving out a substantial doctrinal role for LWD in policing claim scope, separate and distinct from the enablement requirement. Another example can be found in In re Kubin, another appeal to the Federal Circuit of a Board decision finding a claim enabled but nevertheless in violation of LWD due to the excessive breadth of the claim. The Federal Circuit's decision in Alonso does not bode well for Kubin, but the facts are distinct and we will have to await the Federal Circuit's decision in that case.
The PTO routinely allows claims broadly reciting a genus comprising any isolated antibody capable of recognizing a specified antigen, without limiting the applicant to any specific antibody, or any specific epitope on the antigen (a single antigen typically has multiple epitopes, the specific and discrete region of the antigen bound by a given antibody). The PTO's revised written description training materials, published earlier this year, specifically states that the disclosure of a reasonably well characterized antigen satisfies the written description requirement with respect to a claim encompassing any isolated antibody capable of binding the antigen (see earlier post discussing the training materials). This is consistent with the Federal Circuit's decision in Noelle v. Lederman, which found that although the disclosure of a mouse antigen was insufficient satisfy the written description requirement with respect to a claim reciting a monoclonal antibody capable of recognizing the human counterpart of the mouse antigen, disclosure of the mouse antigen would provide adequate written description support for a claim broadly reciting any antibody capable of recognizing the mouse antigen. It is worth pointing out that under this approach, not only is it unnecessary for the patent applicant to disclose the structure of any monoclonal antibody falling within the scope of the claim, it is also unnecessary to disclose any information regarding the structure of the antigen. This is inconsistent with other Federal Circuit written description cases relating to gene sequences which have required some disclosure of chemical structure, i.e., Fiers v. Revel, UC v. Lilly, In re Wallach, and most recently Carnegie-Mellon v. Roche (discussed in an earlier blog post).
Alonso’s claim is distinguishable, however, because it is not directed to a single antigen. The PTO and Court found that this sort of tumor is characterized by “considerable antigenic heterogeneity,” both between patients and between metastatic sites within a single patient. It was this unpredictability with respect to the structure of neurofibrosarcoma antigens, and consequently the genus of antibodies recited in the claim, which was the basis for the written description rejection.
In Alonso, the PTO is employing the written description requirement as a doctrinal tool for limiting claim scope. Traditionally, the enablement requirement has played the primary role in policing claim scope. For example, in Amgen v. Chugai the enablement requirement was invoked to invalidate a claim broadly reciting genetic sequences encoding structurally undefined functional analogs of human erythropoietin, with the court essentially finding that the claimed genus exceeded the scope justified by Amgen's disclosure of a single gene encoding wild type erythropoietin. But in UC v. Lilly the Federal Circuit established a new role for the written description requirement, which I refer to as the Lilly written description requirement (LWD), and which many commentators have characterized as a super-enablement requirement. Indeed, the examiner rejected Alonso’s claim for failure to comply with both the enablement and written description requirements, treating the two doctrines as effectively redundant. The Board, however, only affirmed the written description rejection, reversing on the issue of enablement.
The examiners tandem enablement-LWD rejection was not surprising. For years, examiners have asserted violations of both the enablement and LWD requirements in rejecting what they perceive to be overly broad genus claims. But until recently, it was almost unheard of for the Board to explicitly overrule the enablement rejection while affirming on LWD. In fact, when I conducted a comprehensive study of LWD cases a couple of years ago I found only one case where the Board (or any court for that matter) reversed an enablement rejection but affirmed on LWD. That decision, Capon v. Esshar, was soundly reversed by the Federal Circuit on appeal. In its decision, the Federal Circuit stated that analysis for compliance with the enablement and LWD requirements are essentially coextensive, and suggested that it would rarely if ever be proper to find a claim enabled but nevertheless in violation of LWD. Thus, in this respect Alonso breaks new ground.
The Federal Circuit has repeatedly stated that an applicant can satisfy LWD with respect to a genus claim by disclosing "a representative number of species in that genus." In Alonso the court agreed with the board that, in view of the heterogeneity and unpredictability of the antigen, disclosure of a single species did not constitute a “representative number." However, the decision provides no guidance with respect to how many species would represent a representative number. In fact, although the "representative number" standard was set forth in UC v. Lilly in 1997, a cited repeatedly by the courts and PTO, to my knowledge neither the PTO nor the courts have yet to provide any useful guidance with respect to how to determine just how many species is necessary to constitute a "representative number.” In fact, I would argue that the test for compliance of a genus claim with LWD cannot be meaningfully distinguished from the test for enablement, as suggested by various panels of the Federal Circuit, e.g., Capon v. Esshar and Lizardtech.
Alonso exemplifies what I see as a recent emphasis in the patent office on carving out a substantial doctrinal role for LWD in policing claim scope, separate and distinct from the enablement requirement. Another example can be found in In re Kubin, another appeal to the Federal Circuit of a Board decision finding a claim enabled but nevertheless in violation of LWD due to the excessive breadth of the claim. The Federal Circuit's decision in Alonso does not bode well for Kubin, but the facts are distinct and we will have to await the Federal Circuit's decision in that case.
Subscribe to:
Posts (Atom)
Posts
