Taking it personally: patents, medicines and genetic markers

Purr-sonalised medicine
The IPKat is all in favour of medicine which, he thinks, can come in really handy -- even when you're unwell.  He was therefore fascinated to learn of a recent discussion in Europe of a topic which has already exercised some of the finest minds of his American colleagues: the patenting of inventions relating to personalised medicine.  Having heard about this from katfriend Suleman Ali (Holly IP), he is pleased to say that the latter was willingly persuaded to write a short note on the topic for the benefit of readers of this weblog. Here it is:
"Is the EPO changing Its stance on personalised medicine inventions?

Case law is an important means by which we know what is patentable at the European Patent Office (EPO). However, sometimes the EPO’s view of what is patentable in an area changes before the case law does. This can sometimes be detected when Examiners start raising objections they would not have previously done. Clearly, applicants need to know about such changes as soon as possible so that they can revise their filing strategies and re-evaluate their expectations of the claims they are likely obtain. Meetings between the EPO and the epi (the professional institute for EPO attorneys) are very useful forums for obtaining ‘inside information’ about the EPO’s thinking which is not yet apparent from the case law. The June 2012 issue of epi Information provides a report of such a meeting held on 10 November 2011 between the EPO and the biotech committee of the epi. Discussion item 8 is reported as follows:
‘8. Inventions in the area of pharmacogenomics 
This concerns cases which are based on a genetic marker to treat a disease, for example methylation profiles. It can involve a new patient group defined by an SNP. The EPO said that often the claims can lack novelty, as one patient will have inevitably been treated with the SNP, even if the art does not explicitly say so.’
The EPO’s comments seem to indicate that it is about to change the way it assesses novelty when looking at medical use claims that refer to treatment of a specific patient group.

To give a little technical background to the EPO’s comments, an SNP is a form of genetic marker which varies between individuals. The idea behind the relatively new field of pharmacogenomics is that, if you know which SNP variants a patient possesses, you can personalise the drugs given to a patient in accordance with his genetic makeup. It is now recognised that the genetic makeup of an individual can be very influential as to whether he responds to a drug, and so one application of pharmacogenomics is to only give those drugs to patients who will respond to them.

Personalised medicines can also be based on non-genetic biomarkers, such as the level of virus the individual has.

Personalised medicines offer the potential to use drugs much more effectively. That is clearly of benefit to patients, but should also help to reduce costs in times when many governments feel increasingly dismayed at the yearly increases needed to health budgets. The sector most likely to benefit in the short time is cancer therapy where most of the work in identifying biomarkers is focussed. However, biomarkers are increasingly being sought for many other diseases.

Presently, suitable biomarkers for personalised medicine are proving difficult to find. So it seems that the sector is going to require a lot of investment -- but in investors in biotech do like to see that strong patent protection is available in the relevant sector.

Personalised medicines, and in fact diagnostics in general, has been thrown into uncertainty in the US after the Supreme Court’s decision in Mayo v Prometheus [on which see earlier Katposts here and here] which found that a claim referring to steps that determined the level of a drug in a patient was directed to a law of nature and was thus not patentable. It would be unfortunate for personalised medicines to be dealt a further blow by the EPO, making the test for novelty stricter in this area.

Claims for personalised medicine inventions can have many different forms, but typically they are along the following lines:
Substance X for use in a method of treating condition Y in an individual with biomarker Z’.
There is an argument here that perhaps applicants only deserve claims to the method of selecting the individual (by detection of the biomarker), and not to treatment of the individual. However there is a lot more money in therapy, with figures being quoted of 6% versus 94% for the money to be made in selection versus therapy. Since personalised medicine results in therapy being more effective, there is an argument that the applicant deserves claims to the therapy step.
The crux of the present issue is whether limiting a medical use claim by specifying that the individual has biomarker Z will confer novelty where the prior art is silent about patients having biomarker Z, but where patients with biomarker Z will inevitably have been treated, i.e. does limiting a medical use claim to a patient group that overlaps with, or is within, the prior art patient group, make the claim novel?

The earliest case to tackle the issue seems to have been T233/96 which gave a strict two-part test for novelty requiring the patient groups to be non-overlapping and for there to be a functional relationship between the biomarker and the therapy, i.e. the patient group could not be an arbitrary group. However, subsequent case law has not followed the test. In T1399/04 the Board cited T233/96, but took a different view, generously allowing claims which covered more than 50% of a prior art patient group. Decisions T836/01 and T1642/06 also allowed claims where patient groups overlapped with the prior art.

Based on the comments at the EPO/epi meeting and from the experiences of attorneys I know who are handling European patent applications in this area, it seems that EPO is taking a stricter view of the issue, and is probably looking for a test case to change the case law. If the EPO decides on a test which is based on the concept of a patient with the relevant biomarker ‘inevitably’ having been treated, presumably this is a prior use test, in which case it would be burdensome for applicants to locate evidence on what actually happened. However if the test is similar to that used in T233/96, i.e. requiring that patient groups do not overlap, then it will have the effect of severely curtailing patent protection for personalised medicines because most drugs are initially given to everyone with the condition.

I hope that the EPO will be wise enough to recognise that making the test for novelty stricter for medical use claims limited by patient group will have a substantial impact on the patent protection that can be obtained in the area of personalised medicines, at a time when this very promising sector needs all the support it can get".
Thanks, Suleman, for this most instructive piece, says the IPKat.  Merpel is fascinated by this for quite another reason, though.  It reflects a growing trend towards what might be termed "mass personalisation".  We have it in branding and marketing, where the use of sophisticated software in reading your emails and online purchases enables a personalised dose of advertising to be specifically targeted at the individual. It also exists in the design and fashion sector, where a combination of interactive software and manufacturing improvements produces the result that a purchaser of, say, sports shoes, can determine the style, size, colour and bolt-on features that characterise it, rather than going into some random shop and putting a tentative foot into a sample shoe that might previously have been tried by someone with sweaty socks and fungal growths between the toes ...

Your own personalised medicine here and here [not for the squeamish]
Taking it personally: patents, medicines and genetic markers Taking it personally: patents, medicines and genetic markers Reviewed by Jeremy on Friday, August 17, 2012 Rating: 5


  1. I can see the novelty problem with "Substance X for use in a method of treating condition Y in an individual with biomarker Z" even if the biomarker is previously (a) not known or (b) not recognized as affecting the therapy, but is it perhaps avoidable by a minor modification of the claim to "Substance X for use in a method of treating condition Y in an individual known to have biomarker Z"?

  2. In response to Derek's comment: your point is very interesting. Similar strategies involve defining the patient as one who 'has been selected on the basis of possessing biomarker Z' or one that 'has been tested for possessing biomarker Z wherein the test was positive'. However the problem with this approach is that Examiners can say that knowledge of the patients to be treated does not result in a 'new patient group' because structurally they are the same as the prior art group in respect of possessing features relevant to the claim. Testing them for possessing biomarker Z did not give them a physical feature that can be used to define a new patient group. However of course on a good day, an Examiner might let such a claim through, so it's worth trying.

  3. I think that the underlying problem here is that the use of biomarkers is a therapeutical method, which has been explicitly excluded by the European lawmaker from patentability. To an outsider, all this talk about "personalized medicine" looks like a rather transparent attempt to undermine this exclusion, using an exception introduced in the examination of novelty which was specifically introduced to allow the patentability of known substances for second medical uses, without allowing medical methods themselves.

  4. If a patient with the biomarker has been treated with the medicine and their access to the medicine was not under a duty of confidence then the claim as proposed (including that proposed by Derek) lacks novelty.

  5. @Suleman:
    Thank you for your response, but I don't understand the reasoning (not yours) of "structurally they are the same as the prior art group in respect of possessing features relevant to the claim". Surely the point of biomarker testing is to find people who benefit from a particular therapy, as HER2 and Herceptin, and the group of people who benefit from Herceptin (those who are HER2-positive) is distinct from the group of those who do not (those who are HER2-negative); and that feature (known biomarker presence) is relevant to the claim because the claim says in effect "treat those who will benefit, not those who won't". But I have yet to read the cases you cite, and perhaps I will find enlightenment there.
    @Anonymous (18 August, 10:32:00 PM):
    My proposed claim does not lack novelty merely because a patient with the biomarker has been treated, as long as the patient is not known to have the biomarker, nor do Suleman's alternatives for similar reasons. Possessing the biomarker is a characteristic of the patient regardless of knowledge (hence the novelty problem mentioned by Suleman in his original note if the treatment has already been given without testing for the biomarker), being known to possess it requires something more and that is where the novelty is created.

  6. In response to Derek again: your arguments might work in getting a case through. My previous response to you essentially describes ways in which Examiners might disagree, i.e. saying that whether or not a patient is 'known' to have a biomarker is not a technical feature that will give novelty. I would say the question is very much an open one. I suspect in the real world some Examiners may agree with you and some may not.

  7. To Freyberg:

    I would say, in the case of herceptin it is different because HER2 is not just a marker: HER2 is the target of the drug (herceptin), thus there is a direct technical relationship between the presence of HER2 and the efficacy of herceptin. You know that the drug would not work if the patient did not express HER2 thus you would not treat her with that drug. But when it's just "a" molecular marker like an SNP, you don't know.

    In many cases, the presence of some genomic markers like SNP's just CORRELATES with some feature of the patient (e.g. a particular sensitivity to a drug) without presence of a true CAUSAL relationship between the presence of the marker (SNP) and the sensitivity to the drug. There will always be some patients who have the marker but who nevertheless will be neither more nor less sensitive than other patients to the drug (because the correlation will not be 100%) and vice versa. I would say, in that case it is still the same medical use (thus not novel) because you're trying to treat the same disease using the same means, and based on the same molecular mechanism (technical effect).

    Imagine that a doctor realizes that a drug works better with patients whose name starts with the letter "A". Does that make the use of the drug new on these patients ? Of course not. It's just treating the same patients for the same disease with the same drug.

  8. "The earliest case to tackle the issue seems to have been T233/96 which gave a strict two-part test for novelty requiring the patient groups to be non-overlapping and for there to be a functional relationship between the biomarker and the therapy, i.e. the patient group could not be an arbitrary group. However, subsequent case law has not followed the test. In T1399/04 the Board cited T233/96, but took a different view, generously allowing claims which covered more than 50% of a prior art patient group. Decisions T836/01 and T1642/06 also allowed claims where patient groups overlapped with the prior art."

    Well obviously you haven't read T1399/04 carefully because it does not say that T233/96 is wrong, it says that "the present case differs obviously from the situation underlying decision T 233/96." In T836, novelty was recognized because "the claimed invention relies upon a different technical effect from the one disclosed in document (1)" thus novelty had nothing to do with the group of patients. Same in T1642.

    Again, patent attorneys who do not get their ways invent a fake controversy, allege that the EPO's examiners and boards are not consistent and claim that the cure is to grant everything they want.

  9. It is true that T1399/04 identified the factual situation to be different to T233/96. However, T1399/04 also disagreed with the way T233/96 interpreted T19/86 and T893/90 in formulating the two-part test (see point 35 of Reasons for the Decision "The present Board does not see basis for this interpretation in the relevant parts of decisions T19/86 (points (5) to (8) and T893/90 (points (4.2) to (4.6))". Therefore T1399/04 did disagree with the rationale applied by T233/96.

    Also, whilst T836/01 and T1642/06 do indeed involve cases where the claimed invention relies upon a different technical effect from the prior art, the point still remains that the claimed patient groups overlapped with the prior art patient groups. Contrary to T233/96 which requires the claimed patient group to be non-overlapping with the prior art, T1642/06 states that "The overlap in the therapeutic application of the use of the prior art and the use of the claim is irrelevant, because the technical effect stated in the claim identifies a new clinical situation and remains different from that of the prior art" (point 2.1.2, 3rd para, of Reasons for the Decision). Thus the later cases do seem to be taking a different approach to the strict rationale used in T233/96, particularly with regard to the requirement for non-overlapping patient groups.

  10. Let's imagine pathogens A and B. Some people are infected by A, some by B, some by both A and B. X is a well-known drug against A. A doctor finds that the drug works better (e.g. has less side effects) on patients who are infected with A and B than on patients who are infected only on A, and he finds that it is because there's a component in B that, when contacted with X, causes an effect on pathogen A which is beneficial to the patient.

    If somebody claims "use of X to treat patients infected with A and B", of course that overlaps with the group of patients "infected with A". But because there is a new TECHNICAL EFFECT, the use is new.

    Now if there was no technical effect i.e. X works just as well on patients infected with A as on those infected with A and B, then claiming X for treating patients infected with A and B creates a new group of patients, but one that overlaps with the generic group (patients having A) but because there is no technical effect this new group does not lead to novelty (because it's an arbitrary group). That's what the decisions mean, and if you read them well, you'll see that they all agree on that.

  11. Paras 22 to 29 of T734/12 has interesting comments on patient groups

  12. Subsequent decisions T108/09 and T734/12 show the Boards of Appeal are taking a lenient approach

  13. epi information June 2014 refers to the EPO mentioning T734/12 at a meeting with the epi on 20 November 2013, so clearly this is a case which the EPO wants to highlight for its approach on novelty of patient groups.


All comments must be moderated by a member of the IPKat team before they appear on the blog. Comments will not be allowed if the contravene the IPKat policy that readers' comments should not be obscene or defamatory; they should not consist of ad hominem attacks on members of the blog team or other comment-posters and they should make a constructive contribution to the discussion of the post on which they purport to comment.

It is also the IPKat policy that comments should not be made completely anonymously, and users should use a consistent name or pseudonym (which should not itself be defamatory or obscene, or that of another real person), either in the "identity" field, or at the beginning of the comment. Current practice is to, however, allow a limited number of comments that contravene this policy, provided that the comment has a high degree of relevance and the comment chain does not become too difficult to follow.

Learn more here: http://ipkitten.blogspot.com/p/want-to-complain.html

Powered by Blogger.