“Personalized medicine,” in its broadest aspects, is essentially the tailoring of treatments to individual characteristics, needs and preferences. More commonly, the term is used to refer to the tailoring of treatments to an individual based on his or her personal genetic makeup and prior treatment history.

The development of personalized medicine approaches involves much time, effort and expense, and there is a strong interest in protecting innovations in this area in a meaningful way. In Europe and Canada, methods of medical treatment are not considered to be patentable subject-matter, with both jurisdictions taking the approach that patents ought not to interfere with the ability of physicians to exercise their skill and judgment. Protection for therapeutic methods can nevertheless still be obtained in both jurisdictions by drafting claims in an acceptable “use” format.

Treatment-related personalized medicine claims are by definition “second medical use” claims, in that they relate to the use of a known therapeutic. An added complication is that such claims typically also relate to the use of that known therapeutic to treat the same disease as previously treated, differing only in one or more or the population being treated, the route of administration, the dosage amount and/or the dosage regimen.

In the following, we summarise briefly the approaches being taken by the courts and the patent offices in Europe and Canada with respect to such claims.


In Europe, methods of medical treatment are specifically excluded from patentability by statute (Article 53(c) EPC, which also allows for use-limited product claims).[1] Article 54(5) EPC, introduced with the EPC 2000 revisions, allows for claims to second medical uses.[2]

Prior to introduction of Article 54(5) EPC, the patentability of second medical uses had been established in the Enlarged Board of Appeal Decision G5/83. Several Technical Board of Appeal decisions considered the issue of what constituted a “specified new and inventive therapeutic application” in accordance with G5/83, particularly with respect to treatment of sub-populations. In one of the earliest and frequently referenced decisions, T19/86, which related to the claimed use of a known therapeutic for prophylactic treatment of the same disease in an immunologically different population of animals, the board concluded that G5/83 should be broadly construed. In particular, the board indicated that the question of whether a new therapeutic use is in accordance with G5/83 should be answered not only on the basis of the ailment to be treated, but also on the basis of the subject or patient to be treated.

In T19/86, the prior art and claimed populations were distinct (sero-negative versus sero-positive piglets), however, in many instances a claimed sub-population may overlap to some extent with the previously treated population. The issue of overlapping populations was considered in T0233/96, where the board interpreted preceding decisions as indicating that in order for a use of a known treatment in a patient sub-group to be novel, the sub-group could not overlap with the group previously treated. In addition, the sub-group must be distinguishable from the prior population by its physiological or pathological status, and selected based on a functional relationship between this physiological or pathological status and the therapeutic effect obtained (rather than being an arbitrary selection).

The board in T1399/04 also considered overlapping populations of patients, but in this case found that despite an overlap in treated populations, the claimed use was novel because the recited sub-group was defined by a previously undisclosed pathological and physiological status, and its selection was not arbitrary. In particular, the board focused on the demonstrated functional relationship between the claimed sub-group and the improved effect of the claimed treatment. In a more recent decision, T0734/12, the board likewise found claims directed to a known treatment using a particular dosage regimen in a specified sub-population of patients to be novel, even though the dosage regimen was known for the same treatment of a different but overlapping population. The board held that the claimed sub-population had a different physiological and pathological status to the prior population, thus rendering the claimed use novel. The use was, however, found to lack an inventive step, primarily because the patent did not include any data demonstrating an improvement over the prior art treatment.

In contrast to the fairly expansive approach taken in the above decisions, however, the approach being adopted by the European Patent Office (EPO) to claims relating to treatment of sub-populations appears to be more limited. The EPO has expressed the opinion that inventions in the area of pharmacogenomics such as those involving a new patient group defined by a biomarker, often lack novelty.[3] Specifically, the EPO has indicated that in such cases, it assumes at least one patient with the biomarker inevitably has been previously treated, even if the art does not explicitly say so, and the claim is therefore anticipated. Thus, the EPO can be expected to raise novelty objections against a claim directed to use of a known treatment in a sub-population of patients with a specified biomarker, if it can be established that the biomarker is present in a significant proportion of patients.

There is some indication, however, that the EPO may consider a claim drafted as a diagnostic-style “identify and treat” claim reciting an active step of determining whether or not a patient has a particular biomarker and then treating the patient accordingly, to be novel. Novelty in this case resides in the step of determining whether the patient has the biomarker, and will be dependent on whether a link has been established between the presence/absence of the biomarker and an improvement in the treatment.

The inclusion of supporting data in an application is thus emerging as a key factor in obtaining personalized medicine claims in Europe. Such data will be essential to establish that selection of any claimed sub-population/dosage regimen/route of administration is not arbitrary, but based on a functional relationship between the claimed feature and an improved effect of the claimed treatment.


In contrast to Europe, exclusion of methods of medical treatment and patentability of second medical uses in Canada are both based in case law.[4],[5]  Also in contrast to Europe, where claims that are drafted in the correct “use” format are assumed to be directed to patentable subject-matter,[6] care needs to be given to the exact wording used in medical use claims in Canada to ensure that an otherwise correctly formatted claim cannot be equated to a method of medical treatment. For example, inclusion of a medical or surgical step, or a step or aspect that may be construed as requiring the skill and judgment of a physician,[7] will result in a medical use claim being deemed non-statutory. Many of the challenges encountered when prosecuting and defending personalized medicine claims in Canada stem from the fact the claims are being construed as relating to non-statutory subject matter.

Of note in this regard are three practice notices recently published by the Canadian Intellectual Property Office (CIPO) outlining revisions to the CIPO’s practice for determining whether claimed subject matter is statutory.[8] The most relevant of these practice notices to personalized medicine claims is PN2013-04 “Examination Practice Respecting Medical Uses,” which expands on the guidance provided in the Manual of Patent Office Practice (section 17.02.03). The notice is primarily focused on claims that recite dosage regimens or dosage ranges, but also discusses other aspects of personalized medicine, including treatment of a sub-population of patients.

PN2013-04 follows a purposive construction approach when analyzing the claims to determine whether the claimed subject-matter is statutory, which requires that the “essential elements” of the claim be identified.[9] If an essential element is considered to be a medical or surgical step, or to require a physician’s skill or judgement, then the claim is considered to be directed to a method of medical treatment.[10]

In particular, the CIPO considers that if, purposively construed, the claim relates to “what” to use to treat a patient (for example, a composition, formulation or dosage form), then the claim is generally acceptable. However, if the claim relates to “how” to treat a patient (for example, a route of administration or a specific dosage schedule), then the CIPO considers that the claimed use is non-statutory. In adopting this approach, the CIPO references the Supreme Court decision of Apotex.[11] Interestingly, the use claim considered in Apotex does not refer to “how” or “when” the drug is administered, but rather defines a dosage form, which was determined to be statutory.[12] An alternate interpretation of Apotex would be that if no professional skill is required to determine the “how” or “when,” in a manner analogous to providing a dosage form, then the claim should be acceptable. PN2013-04 appears to simplify the determination of the requirements for patentability by suggesting that any claim directed to providing an invention based on the “how” (or “when”) is not statutory.

PN2013-04 further states that essential elements “that narrow treatment to a patient sub-population (rather than bring treatment to a new population) or administration site,” are also considered to limit a physician’s professional skill or judgment, and thus are non-statutory. Example 7.1 of PN2013-04[13] is provided to illustrate this scenario and provides a sample claim relating to treatment of a “sub-population” of patients who respond more efficiently to a known treatment (“compound X”) due to their genetic makeup. The CIPO indicates that such a use would be considered non-statutory as it restricts the choices of a physician on how to use compound X. Accordingly, claims directed to sub-populations are considered non-statutory under current Canadian practice, even though this determination is not based on any Canadian jurisprudence.

Interestingly, PN2013-04 also indicates that the use in Example 7.1 would not be considered to be a selection because “…in a selection, all of the range claimed must be novel,” whereas some of the patients in the example would previously have been treated using the same drug without knowledge of their genetic makeup. This interpretation by CIPO of the test of a proper selection appears to be controversial and inconsistent with the conclusion arrived at by the Supreme Court of Canada in Sanofi.[14]

The features established in Sanofi as defining a valid selection include an advantage (or disadvantage) shared by the selected group; the whole of the group possesses the advantage, and the selection is in respect of a quality or character peculiar to the group. While Sanofi also established that the compounds of the selection cannot have been made before, it is of interest that in Sanofi the selected compound was a stereoisomer with beneficial properties when compared to the known racemic mixture of D- and L-isomers.  The beneficial isomer had therefore been previously administered to a patient as part of the racemic mixture, however, the court still determined that this selected isomer was not anticipated or obvious and was statutory. 

Nonetheless, CIPO’s current practice, like the EPO’s, is to reject any claims directed to the use of a known therapeutic to treat the same disease in a specified sub-population, such as patients with a specified biomarker. In Europe, these objections will be raised on the basis of lack of novelty, whereas in Canada, the objection will be that the claimed subject-matter is non-statutory. Both positions, however, appear to be at odds with the respective case law.

In Canada also, re-drafting such claims as “identify and treat” claims, which the EPO have indicated may be acceptable, is unlikely to be a viable option, as inclusion of a treatment step would result in the claim being construed as being directed to a non-statutory method of medical treatment. An alternative would be to amend the claim into a pure diagnostic claim. Another option may be to include a diagnostic step in a standard use claim in order to further define the patient being treated, for example, by specifying that the patient has been identified as suitable for treatment by screening the patient for a specific gene mutation.


This overview highlights some of the challenges in obtaining protection for personalized medicine inventions in Europe and Canada. While each jurisdiction is developing its own unique approach as to how to personalized medicine innovations should be protected, one commonality is that respective approaches currently being adopted by the patent offices are not necessarily in line with the relevant case law. Given that innovator companies are continuing to invest large amounts of capital into the personalized medicine space and to file patent applications for their innovations, it appears inevitable that the current patent office approaches will be challenged. It is to be hoped that more consistent case law and patent office practice will emerge as a result, as it has in more traditional practice areas. In the interim, there appears to be sufficient leeway in the respective patent office practices for applicants to obtain meaningful protection for personalized medicine inventions with some forethought and a creative use of claim language.