Introduction

Issues surrounding gene patents and access to health care services have been a subject of wide debate for quite some time.1, 2, 3, 4, 5, 6, 7, 8 With genetic testing moving into mainstream medicine4, 9, 10 and with ongoing efforts to harmonize and to set standards for genetic testing in Europe and elsewhere,11, 12 patenting and licensing issues regarding genetic testing continue to spur further attention from all parties involved. Opponents still question the legitimacy of patents claiming genes, and further argue that the availability of alternative testing is squashed while costs of testing could increase without guaranteeing good quality. Gene patents are perceived as hampering both innovation and access of patients to adequate genetic testing.

At present, solid empirical data to support statements on the impact of gene patents on diagnostic testing services are still scarce. Previous studies mainly considered gene patent documents based on keyword search strategies, allowing the identification of patents actually claiming genetic material (eg, DNA). However, any interpretation of the real subject matter of the claims in the patent document is often lacking13, 14 and many of the listed patents are often more relevant to the development of therapeutics than to genetic diagnostics.

The exclusive right granted to a patentee entails a right to prevent or exclude others from making, using, selling, offering for sale or importing the claimed invention without authorization of the patentee,15 during a limited period of time and within a certain territory. A patent does not impose an obligation to the patentee to enforce his rights. Rather, a patent offers an opportunity to the patentee to undertake action against infringers, ranging, for example, from sending cease-and-desist letters to starting a lawsuit, or to provide a permission or license to lawfully use the patented invention. The possible curtailing effect of patents on the use of a patented invention by third parties thus largely depends on the degree and way of enforcement of the patent by the patentee, and on the availability and terms of licensing deals. Patents not actively enforced, or patents licensed out under reasonable conditions, seldom cause great concern regarding the use of the patented technology. Recent studies discuss the effect of gene patent enforcement in the United States1, 8 and the attitude of clinical-genetic laboratories2 on access and use in more depth.

The present study reports on patenting and licensing practices related to genetic diagnostics of spinocerebellar ataxias (SCAs), an important group of neurogenetic diseases affecting patients worldwide.16 Within the autosomal, dominant SCAs, about 30 monogenic subtypes are currently distinguished through genetic mapping, with the causative gene being identified for 16 subtypes (Supplementary Table 1). The most accurate way to diagnose this genetic disease is molecular genetic testing, using a gene panel that typically includes individual assays for the five most prevalent SCA subtypes (SCA1, SCA2, SCA3, SCA6, SCA7) and aiming to detect the pathogenic CAG-repeat expansion in one of these genes.17

We developed a patent categorization method to rank European and US patents according to their impact on SCA testing. Our method facilitates the assessment of actual hampering effects of granted patents towards products and methods used in diagnostic testing, and could similarly serve to assess freedom to operate in other fields of technology.

Methods

Patent search

Patents related to SCA genes and genetic diagnostics were extensively searched in both public and commercial patent databases, largely as described earlier18 As claims may change over time (between filing and granting), the patent information provided here was last updated on March 1, 2010.

Legal status of patents and patent applications

The legal status of European and international PCT (Patent Cooperation Treaty) patent documents was checked on the European Patent Offices's website (http://www.epoline.com). The legal status of US patent documents was checked on the US Patent and Trademark Office's website (http://www.uspto.gov) and its Patent Application Information Retrieval website (http://portal.uspto.gov/external/portal/pair).

Claim analysis

Claims were analyzed and interpreted in light of the patent specification. We evaluated the reasonable scope of the pending/granted claims based on the context (eg, the claim itself, other claims, the description, drawings, file wrapper). In Europe, Art. 69 of the European Patent Convention (EPC) and its Protocol form the basis for such interpretation, striving for a balance between a fair protection for the patentee and a reasonable degree of certainty for third parties. In the United States, basis for claim interpretation can be found in the US Utility Patent Act §112, asking for a ‘clear written description’ and the ‘best mode for carrying out the invention’.

Licensing policy

Telephone interviews were made to unravel licensing policies of Athena Diagnostics (Boston, MA, USA), Baylor College of Medicine (Houston, TX, USA) and the Research Development Foundation (Carson City, NV, USA).

Patent thickets

In order to evaluate whether a patent thicket exists or could be emerging in the field of SCA genetic diagnostic testing in Europe or in the United States, all granted patents currently in force and all patent applications at present under examination in Europe and the United States, were listed in this study.

Results

The patent categorization method

A patent categorization method has been developed to weigh the impact of patents on genetic diagnostic services for SCA (Figure 1). A patent is considered to be hampering access to a genetic diagnostic test if the relevant patent (step 1) is valid in the country of interest (step 2), claims an essential part of the genetic test (step 3) that cannot be circumvented (step 4), is actively enforced and is not licensed at a reasonable cost (step 5).

Figure 1
figure 1

The patent categorization method for the assessment of hampering effects of patents related to genetic diagnostics. The metaphors of step 5 are modified from http://GenericsWeb.com.

Full size image

Patent search (step 1)

Keyword searches in the title, abstract, and claims of patents in relevant International Patent Classification (IPC) classes yielded over 100 patent documents per SCA subtype, identifying the respective SCA gene. Despite careful drafting of Boolean keyword strings, the occurrence of semantics as well as the obfuscation of claim language occasionally found in patent documents made it necessary to check each patent document individually for relevance to the study. Manual processing of the patent claims, as opposed to automated patent analyses, remains the most optimal albeit time-consuming approach.

The searches yielded 19 patent families related to SCA genetic testing (Table 1). The assignees thereof are geographically distributed, notably in Europe (FR, NL), the US and Asia (JP, IN, KR), reflecting the worldwide prevalence of SCA among all ethnic groups. Within each patent family, we considered the patent documents filed at the EPO, the USPTO and the World Intellectual Property Organisation (WIPO).

Table 1 Overview of SCA patent families
Full size table

A set of 47 patent documents comprising 25 US, 9 EP and 13 PCT patent documents was selected for further study.

The majority of assignees (80%) are public institutions, at present the sole type of assignee in Europe, US and India. The remaining assignees (20%) are private entities in Japan and Korea (Supplementary Figure 1).

Legal status of patents and patent applications (step 2)

The duration and the territorial scope of patents are two essential elements often overlooked by non-patent practitioners. The relevance of a patent can therefore be largely misjudged. Patents can be withdrawn or lapse during or after prosecution, for example, through non-payment of maintenance fees. Checking whether a patent is in force in the country of interest is therefore a prerequisite to any freedom-to-operate analysis or patent landscaping effort assessing current patent obstacles.

Because of a different interpretation of the ‘unity of invention’ concept in the United States compared with Europe (Art. 82 EPC, 35 U.S.C. §121), product and method claims are usually split over different patents in the United States. As a result, patent families often contain more US patent documents. Therefore the relative number of patents is used to compare European and United States data, meaning that results have been averaged out to count one US patent per patent family (Figure 2).

Figure 2
figure 2

Legal status and relative numbers of patent documents within the SCA patent families, considering one United States patent per patent family. ‘Patent application dead’ means that the application is no longer under examination (abandoned, refused, withdrawn or deemed to be withdrawn). ‘Patent pending’ is a patent application that is currently being examined. ‘Patent dead’ means that a granted patent has either lapsed or been nullified. ‘Patent alive’ is a granted patent that is currently valid.

Full size image

In Europe, only one SCA patent has been granted and is still valid. This patent, EP 1.015.628 entitled ‘Large scale genotyping of disease and a diagnostic test for spinocerebellar ataxia type 6’ and owned by the Research Development Foundation (USA), is related to SCA6 testing (also see below). In the United States, a total of 22 SCA patents have been granted of which 20 are still in force. In patent family 9 related to SCA3, US patent 6.124.100 lapsed due to non-payment of maintenance fees, and in patent family 13 related to SCA6, US patent 6.825.332 expired for the same reason (Table 2, Supplementary Table 2).

Table 2 SCA patent landscape using the patent landscaping method of Figure 1. Status on March 1 2010
Full size table

Differences in patent strategies of applicants and examination approaches of the patent offices in Europe and the United States could explain the current legal status of SCA patent applications.13 In Europe, eight patent applications were filed, which ended up ‘dead’: six patent applications are withdrawn or deemed to be withdrawn and two patent applications have been refused, namely EP 1.311.667 of patent family 7 related to SCA2, and EP 932.677 of patent family 13 related to SCA6 (Figure 2, Table 2, Supplementary Table 2). In the United States, three patent applications were filed, which have not led to a patent (yet): one application is still under examination, namely US 2003-0235841 (A) and two patent applications are abandoned, namely US 2005-0032083 (A) and US 2005-0090658 (A) (Figure 2, Table 2, Supplementary Table 2).

Claim analysis (step 3)

The third step in patent categorization involves a meticulous analysis of the claim scope in order to determine the relevance for genetic testing. The most subjective part of any patent analysis is probably the interpretation of the claims in light of the description of the invention, as neither a strict, literal reading of the claims, nor a loose interpretation of the claims as mere guidelines, is appropriate. Rather, a fair protection for the patentee has to be balanced against a reasonable degree of certainty for third parties (Protocol on the Interpretation of Art. 69 EPC). All pending and granted claims were considered valid as it is not the purpose of this study to judge on the validity of the claims in terms of patentability requirements.19

In most fields of technology claims leave room for alternatives or to ‘design around’ the invention, hence triggering innovation. However, claims in genetic diagnostic testing often appear to be truly blocking because of the unique nature of the genetic code and the unique link between a mutation and an inherited disease.20 Blocking claims are often very broad, not allowing for small modifications in molecular method steps to fall outside the scope of the claims. However, a claim towards an entire gene sequence, carrying a predisposing mutation, does not a priori cover the genetic diagnosis for the presence of the naturally occurring gene. Many claims are often directed towards the cDNA sequence or the genomic sequence of the isolated gene. But neither cDNA nor the entire gene is commonly used in genetic testing. Rather, a portion of a gene is isolated and hybridized (eg, to a microarray), or amplified from a larger template and subsequently analyzed for the presence of a pathogenic mutation.

The scope of a claim to a DNA sequence is prone to various interpretations by different national courts. The implementation of the Biotechnology Directive (98/44/EC) resulted in national laws specifying the scope of patents related to biotechnological inventions. However, the absence of relevant case law in most countries leaves the interpretation of these provisions unclear. Moreover, important differences may occur in claim interpretation between United States courts and European national courts.

The scope of claims towards a mutation-carrying gene fragment or preferentially used primers is usually more straightforward. These claims are likely to cause a blocking situation because they protect either a key element or a preferred tool of a genetic test. A license may thus be required. When patented primers are commercially sold, a license is generally provided upon purchase. To circumvent this type of claims, a truly different methodology may be necessary, for example, a non-molecular method (such as protein-based testing). Finding out whether a DNA claim is relevant for a particular genetic testing method cannot be generalized but requires an in-depth analysis on a case-by-case basis.19 Within the SCA patents identified in this study, nucleotide sequences are claimed in 66% of the 47 patent documents. The entire, human SCA genes are claimed in 45% of the patent documents.

A claim to a diagnostic testing method is likely to hamper freedom to operate if it links a mutation to a disease without specifying how to establish this link. Among the (granted and pending) SCA patents investigated, genetic diagnostic testing methods are claimed most frequently (82%), followed by nucleotide sequences (50%), recombinant technology items such as vectors and host cells (41%) and diagnostic kits (27%) (Supplementary Figure 2). Protection for peptides and antibodies, and their use in a diagnostic context is also sought albeit to a lesser extent (Supplementary Figure 2). Accordingly, claims to diagnostic testing methods are not only most commonly envisaged by the applicants, they also appear to be relatively easy to patent. Several types of claims are discussed in further detail.

Claims to human genes

In case the invention comprises the identification of a human gene and the occurrence of predisposing mutations therein, the entire gene is likely to be claimed, and the sequence is often specified. Alternatively, a claim towards a short gene sequence, if open ended, can be interpreted to cover a polynucleotide of any length comprising the claimed sequence, including the entire gene. Claim 10 of US 5.834.183 claims a nucleic acid fragment, defined by the protein it encodes, which could encompass the entire SCA1 gene. Moreover, due to the degenerative character of the genetic code, all sequences that encode the specified SCA1 protein are protected by this claim.

Genetic testing based on PCR amplification of the CAG repeat region is unlikely to infringe this claim because the claimed fragment is neither used nor produced. Firstly, the claimed fragment is unlikely to be used as a PCR template. Secondly, the claimed fragment is more than 580 bp in size and it is unlikely that a PCR amplicon of that size is generated. However, it should be noted that claim interpretation is subject to high variability across various jurisdictions. In the United States, a broader interpretation in the light of the US's strong commercial focus is typically expected. In addition, other claims in this patent may probably infringe PCR methods.

Blocking claims

Claims on the gene sequence, on a common pathogenic mutation, or on the fundamental method to determine the association between a gene and an inherited disease could be blocking in the sense that it would be very difficult to find an alternative product or to invent around the method. In Europe, the single granted SCA patent, EP 1.015.628 of Research Development Foundation (RDF, USA), appears to have a blocking claim affecting SCA6 diagnosis. This claim is drawn towards a method in which the CAG repeat length is first determined and subsequently assessed on its pathogenesis. By claiming the fundamental method of the SCA6 genetic test, it appears to be very difficult to circumvent the patent when performing SCA6 genetic diagnosis. Filed in 1998 as a PCT application, the patent was eventually granted after an examination procedure of about 8 years and validated in 16 countries. No opposition was filed. Meanwhile the patent has lapsed due to non-payment of maintenance fees in Austria, Denmark, Spain, Finland, Greece, Ireland, Italy, the Netherlands, Portugal and Sweden. It currently provides protection in six countries (Belgium, Switzerland, Germany, France, UK and Lichtenstein) until 2018 if maintenance fees continue to be paid.

With SCA6 being part of the common SCA gene panel (including SCA1, SCA2, SCA3, SCA6, SCA7), EP 1.015.628 forbids third parties to execute comprehensive genetic diagnostic testing of SCA in Europe. Whether European genetic diagnostic laboratories could legitimately offer SCA6 testing thus largely depends on the patentee's licensing policy (see ‘Licensing policies’ below).

‘Hidden’ claims

The scope of a patent is not always clear to the untrained eye. Genes or diseases might not be named explicitly but may nevertheless be covered by the wordings of the claims. Identifying these patents by keyword searches is therefore challenging. A patent family entitled ‘A gene related to migraine in man’ and assigned to Rijksuniversiteit Leiden (NL), is, contrary to what the title suggests, drawn to three different diseases related to the CACNA1A gene, including SCA6. Gene sequences and detection methods are claimed, but whether the claims cover SCA6 diagnostic testing is unclear at first sight. Originally, mutations in CACNA1A were linked to familial hemiplegic migraine (FHM) and episodic ataxia type 2 (EA-2). In 1997, the link between a CAG repeat expansion in CACNA1A and SCA6 was published.21 The initial European patent application, EP 834.561 filed in 1996, fails to mention SCA6. A PCT application, WO 9.813.490, was subsequently filed in 1997 mentioning the three diseases related to this gene, including SCA6. The related European patent application, EP 932.677, has meanwhile been refused; the related US patent, US 6.825.332, is expired, and the related US patent application, US 2005-0090658, is abandoned (Table 2, Supplementary Table 2). Remarkable, however, is that while the title and most of the description of the applications EP 834.561 and EP 932.677 appear to relate to migraine and/or episodic ataxia type 2, their claims may also be relevant to SCA6 diagnosis. Recently, the SACGHS case studies have revealed similar situations concerning the patent on hearing impairment and Alzheimer's disease.1, 8

Licensing policies in the SCA patent landscape (step 4)

In cases where genetic testing makes use of a patented product or method, in principle a license has to be negotiated with the patent owner. An exception to such an obligation is the research or experimental use exemption. However, in Europe as well as in the United States, this exemption appears to be very narrow in scope of application. Despite differences in the national research exemption laws in Europe, it is doubtful that it would apply to diagnostic testing.22

Based on inquiries with the patent holders and/or licensees, it appears that most SCA patents are licensed under exclusive terms. In the United States, Canada and Japan, Athena Diagnostics has exclusively licensed-in the patents related to the genetic testing of SCA 1, 2, 3, 5, 6, 7, 8, 13 and 14. Athena pays royalties to the licensor institutions and actively enforces its rights by continuously monitoring and discouraging infringement (Athena Diagnostics, personal communication). Only US 6.855.497 related to SCA10 is currently licensed-out non-exclusively (Baylor College of Medicine, personal communication).

Despite numerous patent applications filed in Europe, only the SCA6 patent (EP 1.015.628) has currently been granted (see ‘Blocking claims’). Clayton Biotechnologies Inc. (Houston, TX, USA) the for-profit company that seeks to commercialize the SCA6 patent on behalf of its patentee (RDF) sent a proposal for non-exclusive licensing to many diagnostic laboratories in Europe on 24 February 2009, even to labs in countries where no patent rights exist (anymore) or where the patents lapsed. The proposed license included an upfront fee of 10.000 USD and an earned royalty of 4% of net sales from all European sales under licensed assay with a minimum annual royalty of 10.000 USD creditable against earned royalties. As far as we know, most labs have not accepted this offer, most probably because valid patents did not (no longer) exist, or because patents did exist, but clinicians contemplated to send samples to labs in one of the countries where no rights exist (anymore). As Clayton Biotechnologies Inc. is after all willing to license-out non-exclusively in Europe, access to the blocking SCA6 patent is principally possible. We therefore categorize this patent not (yet) as a roadblock but as a hurdle that might be overcome in the event of non-exclusive licensing, provided it is affordable to the laboratories.

Assessing impact (step 5)

The data obtained in the previous steps – finding relevant patents (step 1), checking patent validity (step 2), determining scope of claims (step 3) and checking availability of a license (step 4) – have been ultimately combined to systematically rank the patents into five categories according to their actual impact on access to genetic diagnostic testing. Metaphorically speaking, we tend to believe that the impact of patents in the area of diagnostic testing increases from ‘no hampering effect’, to ‘stumbling rock’, ‘hurdle’, ‘roadblock’ or ‘patent thicket’ (Figure 1, Supplementary Information). The patent landscaping method was similarly applied to the patent applications identified in this study. In the absence of details on licensing, we assumed a licensing policy that corresponds to the applicants’ earlier licensing policies or to what was disclosed in personal communications with applicants or licensees.

Stumbling rock

Patents or patent applications in this category claim non-essential elements of a genetic test or elements, which are not related to genetic testing. Access to the technology covered by these patents is rated as not so difficult. The claimed subject matter could for example be exchanged with a known or evident alternative (for example, claims to a very specific method, a specific DNA construct, a vector) or the claim could relate to a rare mutation, which does not represent an essential part of a basic diagnostic screen, such that the claims are not insurmountable for diagnostic testing. Alternatively, licenses could be easily obtainable (for example, when offered at a reasonable royalty to every requester).

The US patent 6.515.197 forms an example of a stumbling rock patent (Table 2). It claims DNA constructs of an ataxin-2-encoding sequence, linked to either a Purkinje cell-specific promoter or a green fluorescent protein-encoding sequence, or both. One of the aims of this invention is the creation of a transgenic animal expressing human SCA2. The method claims in this patent family are all related to the identification of therapeutic agents for treatment of SCA2.

Hurdle

Getting access to patented technology categorized as a ‘hurdle’ is considered to be less obvious. Hurdle patents preferentially protect established methods or primers for which an alternative may not easily be found. Additionally or alternatively, obtaining a license to the hurdle is difficult, for example, because of the cost involved, because the technology is already licensed exclusively to another party and no sublicenses are available, because there is a lack of interest with the patentee for dealing with small-size players, or because of a lack of legal expertise to engage in licensing negotiations – a situation not unusual for a genetic testing laboratory. The US patent 6.623.927 is classified as a hurdle because it claims preferential primers located up and downstream of the CAG repeat expansion (Table 2).

Roadblock

Access to patented technology of this category is considered to be seriously blocked because one or several claims are drawn to an essential element of genetic testing and because circumstances are such that a license can not be obtained. When a patent protects subject matter for which an alternative does not exist or is not directly within reach (see ‘Blocking patent’, Supplementary Information), access to the protected technology would only be possible through licensing. However, in the case of a roadblock, the patent owner does not want to provide licenses or a license has already been granted on exclusive terms to a third party.

Patent thicket

A patent thicket is defined here as a dense web of multiple (blocking) patents held by multiple patent owners.23, 24 Access to the patented technology as a whole is almost completely blocked. Licenses are the only possible solution.

With 20 granted, valid United States patents covering genetic testing of 10 SCA subtypes (Figure 3, Table 2), the situation in the United States can be considered a ‘patent thicket’ as defined above. Exclusive licensing-out of the US patents has placed the rights to genetic testing of SCA in the United States in the hands of Athena Diagnostics. Several labs that infringed Athena's patent rights have reportedly stopped performing genetic testing of SCA, as a response to threats of patent enforcement.5, 25 Thus, Athena has cut through the thicket and is legitimately providing tests to the market, paying royalties to the licensor, but the exclusivity of these services has, in practical terms, given the company quite a monopolistic position, as the freedom to obtain a second medical opinion from an independent laboratory is likely to be restricted.

Figure 3
figure 3

The SCA patent landscape considering granted and pending patents in Europe and in the United States (One Unites States patent is considered per patent family).

Full size image

In Europe, the SCA patent landscape looks much less dense (Figure 3, Table 2). A considerable number of applications were filed but later withdrawn or refused. Reasons for this may have to be sought in the reputedly strict and long examination procedure at the European Patent Office. Yet, even in the absence of a patent thicket, the European situation is potentially problematic because of one blocking patent (see ‘Licensing policies’ above).

Discussion

Stacking of patents may affect both genetic testing laboratories, by seriously reducing a clinical diagnostician's freedom to operate, and patients, by reducing the availability of genetic tests. However, not only patent thickets but also singular patents can throw up a roadblock in the field of molecular diagnostics. Only a few gene patents have direct impact on diagnostic testing, but their scope of protection and exclusive manner of out-licensing is a point of concern.

Our study highlights the striking differences between patenting for SCA in Europe and the United States. The abundance of blocking patents, particularly in the United States, claiming either the gene sequence or the basic method of genetic testing is remarkable. In combination with a restrictive licensing regime, these patents appear to exclude all laboratories from offering the genetic SCA test in the US, Canada and Japan until at least 2015, except the exclusive licensee. On the other hand, even in the absence of a patent thicket, the enforcement of one blocking patent can have far-reaching consequences in some European countries.

The present study is intended to serve as an eye-opener to the genetics community. Geneticists are either not aware of patents related to their tests or choose to ignore them.2 This does not provide a sustainable solution in the long run. Especially in the era of establishing standards for genetic testing,11, 12 it appears necessary to raise awareness on patent matters. Moreover, worldwide efforts to harmonize and standardize genetic testing require a way to practice genetic diagnostics legitimately, without violating IP rights, in line with recommendations made by the OECD26 and the ESHG.27

Recent evolutions in the field of molecular diagnostics at the nanoscale (eg, biochips) not only provide for increased sensitivity and earlier disease detection, but also facilitate multiplexed diagnostics both in preventive and predictive medicine.28 In this way, individuals can be screened rapidly, accurately and simultaneously for the presence or the risk of developing different genetic diseases. For example, inherited mutations in neurological disease genes that could lead to overlapping phenotypes could become part of one single test panel based on several disease genes. It has already been suggested to incorporate genetic testing of SCA2 in the genetic screening of autosomal dominant Parkinsonisms, and more recently, it has been proposed that intermediate-length CAG-repeats in the SCA2 (ATXN2) gene might confer genetic risk for amyotrophic lateral sclerosis.29, 30 As cerebellar ataxia has been reported along with Parkinsonism in SCA1, SCA2, SCA3, SCA7 and SCA1716 and phenotypic overlap has also been documented with FXTAS31 and Huntington's disease,32 multiplexing could become more a general or standard practice in future.

The presented data underscore the need for swift and efficient licensing and/or cross-licensing mechanisms for patents on diagnostic technologies. Non-exclusive and collaborative licensing models in diagnostics, such as patent pools or clearinghouses, have been suggested to render gene patents more easily accessible25, 33 as well as compulsory licenses34 to attenuate the potentially hampering effect of patents on health care services.