Personalized Patents

Patent offices across the globe are struggling to decide whether and how patents should be granted. Here, we present some patent guidelines for analytical companies involved in personalized medicine development in Europe and discuss potential challenges.

Once upon a time, it was straightforward to protect new therapeutic compounds using the patent system. Applicants applied for patents claiming new drugs or new medical uses of known drugs, and attempted to obtain regulatory approval to sell their drug for the patented medical use. The patents and any extensions provided by supplementary protection certificates then expired and generics manufacturers were free to compete with the innovator company.

Nowadays, it is common for research or clinical trials involving a therapeutic product to spark further discoveries about the group of patients in which a drug works or in which it works best. Examples of this type of personalized medicine abound in the scientific literature and patients can be defined by genotype, single-nucleotide polymorphism, epigenetic, or protein markers. Such stratification allows therapies to be selectively prescribed to patients who respond best to the treatment while avoiding treating those who do not respond or in whom the treatment produces adverse effects. In some cases, defining the patient group means the difference between clinical trial success and failure, enabling rational design of smaller trials with high success rates in patients defined by one or more biomarkers. Drugs may be approved in conjunction with companion diagnostics (fast becoming known as “theranostics”) to better control identification of patients who will benefit most.

Patenting lab-based methods of stratifying patient groups provides one form of patent protection for such advances, but the burning question for many innovator companies is whether they can extend patent protection for the therapeutic product itself. It’s a vexed question – especially if the drug has already been described for use in an overlapping or similar population of patients compared with the patients newly identified through the personalized approach. And if this subject matter is patentable, then what protection will be available to companies and what are the real world consequences?

Under the European Patent Convention it has long been recognized that finding a further medical use of a known therapeutic agent is patentable over the earlier known medical use. This acknowledgment led to case law allowing “Swiss” form patent claims directed to the use of the therapeutic agent “for the manufacture of a medicament” for treating the newly identified condition.

Over time, a line of increasingly liberal European Patent Office (EPO) Board of Appeal cases developed these ideas to the point where identifying a new class of patient treatable using a known drug or a new clinical situation constituted patentable subject matter (see Table 1). Notably, the Board in decision T1020/03 explicitly recognized that the investment in clinical trials to find new applications of therapies needed the reward of patent protection to justify it on economic grounds.

After the T1020/03 decision was delivered in 2004, the European Patent Convention was revised and one of the changes was to enshrine the basis for medical use claims in the convention. The EPO chose to do this using different claim wording (“composition for use in a method for treatment”) though publicly stated that it was intended to match as closely as possible the scope of protection provided by a Swiss-type claim

Together, the effect of these changes has been subtle, but favorable to European patent applicants in the field of personalized medicine. The new claim format defines a method of treating patients rather than obliquely referring to manufacturing drugs, and can be easily adapted to “test and treat” claims in which results from analyses identify patients as being eligible for treatment. So the change in law and the development of case law can be used to an applicant’s advantage for patent claims directed to personalized medicine by defining markers or other clinical criteria by which patients can be selected for treatment. If the markers define a subset of patients who would not have been treated without the insight obtained through the analytical step, then there is a good chance that the method will be considered new and inventive, allowing a patent to be granted (subject to meeting all the other normal requirements for patentability).

Things get complicated where the analytical step identifies a group of patients that overlaps fully or partially with the group of patients that the drug was already intended to treat. For example, a drug may already have received regulatory approval for treating a certain disease and may have been administered to a diverse population of patients with the disease. Subsequent identification of a biomarker linked with drug response may subsequently allow more intelligent prescription of the drug by distinguishing between patient subgroups who respond best, those who do not respond, and those in whom the treatment produces adverse effects. Companies naturally wish to obtain patents reflecting this advance, including claims directed to the drug for use in treating the disease in a group of patients specifically characterized as having the biomarker linked with good response. The EPO has struggled with the question of whether treatment of the same disease in a newly-defined patient subgroup represents a genuinely new medical use of the drug, or whether such claims are merely the old medical use described in a different way and, therefore, should be rejected for lack of novelty.

Where a drug was previously described for possible use in treating a disease but no patients were actually treated, or where there was some limited use in patients but with a poor or unknown response rate, then use of the drug for treating patients who have a biomarker indicative of good drug response may be considered novel. In this situation, a patent disclosing a new and non-obvious link between a particular biomarker and a good response to the drug may be able to claim the drug for use in treating the disease in patients who have the biomarker. For example:

Composition X for use in a method for treating leukaemia in a patient, wherein the leukaemia is associated with amplification or overexpression of gene Q, the method comprising administering X to the patient.

Recent developments at the EPO also indicate that including an active step of determining a patient’s genotype (or other markers used in personalized medicine) and treating the patient on the basis of the genetic/biomarker result will avoid objections of lack of novelty. The step of testing the patient to determine the presence of the biomarker is new, even if the drug was previously used successfully for treating the same disease in the same type of patients. Claims in the following format may be allowable:

Resveratrol for use in a method for preventing cardiovascular disease in a patient, wherein the patient has genotype Z, and wherein the method comprises identifying that the patient has genotype Z by genotypic analysis and administering resveratrol to the patient.

And so, the age of personalized medicine claims has been ushered into the patent system in Europe.

For innovator companies this matters a great deal, given the economic costs of drug development. It can mean that effective patent term is lengthened if the personalized medicine applications are filed some years after an initial patent covering the product or its first medical use. In many cases, the types of claim that are granted fit in well with the prescribing information that accompanies the therapeutic agent, and indeed the drug approval. This makes it harder for generics companies to escape infringement. Even by taking advantage of “skinny labeling”, which allows the product label of a generic medicine to omit patented medical uses, generics will still be wary of infringing such patents. Although it is generally believed that in the UK at least, sale of a drug is only likely to be found to infringe a medical use patent if the drug is packaged with an insert mentioning the patented use, this has not been tested in the courts and there are concerns that generics companies might still be caught by “indirect” infringement, if it is known that the drug will be used (either on- or off-label) for the patented use. Therefore, the advent of personalized medicine may allow innovator companies to push back the effective “off-patent” date of their therapeutic products.

With increasing numbers of patents being filed for medical uses based on analytical steps such as biomarker determination, epigenetic analysis and genotype profiling of patient populations, we expect personalized medicine to continue to be at the forefront of the drug patenting battleground for years to come.

Simon Kiddle and Hilary van der Hoff are both partners and patent attorneys at Mewburn Ellis LLP, UK (www.mewburn.com)