Personalised medicine – patenting new drugs from old?

October 2013

An industry under pressure

A combination of high research costs and dwindling opportunities for small molecule targets have diminished the number of candidates entering and progressing through the product pipelines of some leading originator pharmaceutical companies. At the other end of the pharmaceutical product life-cycle patents on the active ingredients of many ‘blockbuster’ drugs are expiring. This allows generic challenges to what are often perceived to be weaker secondary patents and thus increased competition in the marketplace.

A further pressure on pharmaceutical pipelines is the need for each new drug candidate to pass a rigorous approvals process before it can be authorised for public use.  Add to this the need to secure a favourable price and reimbursement from public funds where the emphasis is on the need to show that a drug has clear benefits over and above those already on the market. In fact it has been estimated that only 1 out of 10,000 compounds at the start of development actually make it through to a marketed drug; a ratio that only works commercially if that one drug is a blockbuster.

Production pillPharmaceutical companies therefore have a choice: they either continue to struggle with a wasteful and expensive model of drug discovery and development; diversify into other, less strictly regulated and less costly product areas; or, find a way to improve the efficacy of their existing drug candidates and reduce their side effects.

Furthermore, generic companies themselves now form a mature and sophisticated industry, with many companies merging with the brand companies and some moving into products based on their own research & development; including adapting and improving off-patent drugs that can themselves be patented (so-called ‘supergenerics’).

The strategy for improvement of old drugs by originators and supergenerics alike is increasingly being seen as identifying and targeting for treatment only those groups of patients who respond to the drug – often called 'personalised medicine' – as explained below.

What are personalised medicines?

Conventional drugs are used in the treatment of a population of people grouped together according to their symptoms (e.g. people with lung cancer). However, personalised medicine refers to the treatment of sub-populations within such a group, which suffer the symptoms of that disease because they all share a particular biological characteristic – a ‘predictive biomarker’ – such as a particular gene mutation, which is causally linked to the disease.

When taken by the general group of patients with the disease, not all of whom have the biomarker, a potential personalised medicine may appear ineffective, or even cause adverse events. This is because unwanted effects in people without the biomarker can mask the true potential of the drug. Therefore, once the biomarker to which a drug responds is identified, a drug that is on the face of it unpromising, or even responsible for adverse events, can be demonstrated to have enhanced efficacy.

A leading example of a personalised medicine is gefitinib. In December 2004, gefitinib failed to show significant benefits in an overall population of patients with lung cancer in a Phase III clinical study. It could not therefore enter the European market and appeared to be a failure. But, gefitinib made a surprise comeback. The reason is that it was discovered a sub-population of about 10-15% of lung cancer patients having tumours with a mutation in the epidermal growth factor receptor for tyrosine kinase (EGFR-TK) responded particularly well to the drug. This is because gefitinib inhibits the EGFR-TK activity that promotes the growth of certain lung cancer cells. Subsequently, in June 2009, gefinitib was granted marketing authorisation for the treatment of adults with locally advanced or metastatic non-small cell lung cancer who present positive for mutations of EGFR-TK. Shortly afterwards, the drug was recommended by the National Institute for Health and Care Excellence (NICE) in the UK as a first line treatment option.

The protection challenge for personalised medicines

Label medicineThere is a key commercial challenge to the development of personalised medicines. This arises from the fact that it is not only development of the drug active ingredient that is necessary, but also the related biomarker, or biomarkers, for the disease. The latter can be a very complex research project in its own right. This is a challenge for patent protection, because there is the risk that a considerable period of the monopoly on the active ingredient will elapse before the biomarker is identified and the drug is useful. In many cases, such as those where the original patent is not owned by the party developing the personalised medicine, such as a supergeneric, expiry of patent protection on the original active ingredient may be a pre-requisite.

The question then arises, can a new patent be obtained for use of an old drug in a sub-population identified by a new biomarker.

Patenting drugs for sub-populations

Patents only protect inventions that are new and non-obvious. This is an inherent problem for the type of personalised medicines described above, precisely because they are use a drug that it is already known as a treatment for a disease that is already known. On the face of it, this looks like a novelty problem, but the question is whether patent laws on selection inventions could assist. The most recent UK authority on this issue is the Court of Appeal case of Dr Reddy's Laboratories (UK) Limited v Eli Lilly and Company Limited [2009] EWCA Civ 1362.

In Dr Reddy's the patent being attacked for lack of novelty and inventive step protected a single chemical entity, olanzapine. Olanzapine is an antipsychotic agent used in the treatment of schizophrenia. The validity attack was based on an earlier patent claiming a group of antipsychotic compounds, including thienobenzodiazepines. These compounds were claimed in the form of a so-called Markush formula to which numerous functional groups could be substituted to create a group of compounds within the claim numbering 1019. There was also disclosed in the earlier patent a "preferred" class of 86,000 compounds. Olanzepine was one of the 86,000, although it was not identified specifically. Did this earlier disclosure invalidate the olanzapine patent? The answer from the English Court of Appeal was that it did not. The court relied on EPO case law, specifically Hoechst Enantiomers (T 0296/87) for authority that a compound is not anticipated when it is one of an earlier class of disclosed compounds, unless it is disclosed by "individualised description". Whilst the facts of this case do not concern the type of 'new from old' personalised medicine described above, it does, arguably, indicate that a claim to a compound for treatment of a sub-population described by a specific biomarker may escape anticipation.

MalletAs regards inventive step, English law again looks to EPO case law, in particular Agrevo (T 0939/92). This provides that mere arbitrary selection from a known class of compounds is obvious. However, a selection may be justified as inventive by a hitherto unknown technical effect which is caused by those structural features which distinguish the selected compounds from the wider class of known compounds. Again, although Agrevo was not concerned with biomarkers, its facts are arguably analagous to a claim to a treatment for a selection of patients based on the disclosed structural feature of a biomarker, in which the technical effect is improved efficacy.

SPCs on drugs for sub-populations?

A further problem arises in respect of extending protection by a supplementary protection certificate (SPC): some active ingredients may be sufficiently effective first time round to receive a marketing authorisation (MA) for treatment of a general disease population. However, if it is later found that the active ingredient has a greater effect amongst a particular subpopulation linked by a biomarker, can an SPC be obtained to protect the active ingredient for use in the sub-population using a MA granted for that purpose and despite the earlier MA?

To examine whether an SPC might be available for a personalised medicine in these circumstances, it is again necessary to look at European wide case law, this time from the CJEU.  The Neurim Pharmaceuticals case (Case C-130/11) concerned an application for an SPC to protect melatonin for the treatment of insomnia in humans on the basis of the MA granted in 2007 for this use. This is despite the fact that this active ingredient had earlier been authorised in 2001 and used for regulating the seasonal breeding activity of sheep. At first instance, Mr Justice Arnold in the UK Patents Court held that the CJEU authorities were in favour of a strict approach that restricted SPC protection to products on the basis of their first MA only. This meant that no SPC could be granted for a different therapeutic use of melatonin based on a subsequent MA for that use. However, unexpectedly, and without any reference to the authorities on the strict interpretation, the CJEU in Neurim held that SPC protection for a second and subsequent use of a known active ingredient is permissible.  It does not matter that there is a MA for the same product for an earlier use.

The CJEU seemed to have based its decision on a 'teleological approach', which interprets the SPC Regulation (Regulation (EC) No 469/2009) according to the purpose for which it was drafted. According to this argument, the purpose of the SPC Regulation is to protect the investment made in a product by allowing, in certain circumstances, the extension of monopoly protection to the patent.

ChemicalThe effect of Neurim Pharmaceuticals appears to be that SPCs are not necessarily restricted to products per se, as previously thought, but may extend to second medical uses of the same product where these are the result of their own investment. Doubts have been widely expressed about this decision and questions are currently being referred to the CJEU for a preliminary ruling in another case (GlaxoSmithKline Biologicals (Case C-210/13)) that may throw further light on the teleological approach.  But,if Neurim is correct, and its principles are followed by subsequent decisions of the CJEU and in national courts, it may have implications for personalised medicines: if an SPC can be obtained for a new medical use on the basis that to do so protects investment, why shouldn’t an SPC be granted for a drug for treating a sub-population more effectively than was hitherto possible? As the gefitinib story illustrates, such drugs are themselves the result of investment in lengthy research and development.

The law of patents and SPCs has yet to be tested in the context of the type of personalised medicines discussed above; those that target old drugs at newly identified sub-populations sharing a biomarker. However, as explained, there are starting points from which the law might develop in favour of the protection of such drugs. With personalised medicine becoming more important, it should hopefully not be too long before we find out.

This article was first published in Life Science Intellectual Property Review Newsletter in June 2013.

If you have any questions on this article or would like to propose a subject to be addressed by Synapse please contact us.


Paul England

Paul England

Paul is a senior associate and professional support lawyer in the Patents group based in our London office.

"Conventional drugs are used in the treatment of a population of people grouped together according to their symptoms … personalised medicine refers to the treatment of sub-populations within such a group, … because they all share a particular biological characteristic – a ‘predictive biomarker’."

"Patents only protect inventions that are new and non-obvious. This is an inherent problem for the type of personalised medicines … using a drug that it is already known as a treatment for a disease that is already known. On the face of it, this looks like a novelty problem."