Skip to content

Patenting Medical Diagnostics

The Supreme Court heard oral argument in Mayo v. Prometheus Labs. The case will hopefully provide some guidance on the patenting of medical diagnostics, but because the patent suffers from some real drawbacks, I’m not so sure.  I’ll explain why below. If you are interested in more detail, my 2008 article “Everything is Patentable” discusses the issues in depth.

The types of patents we are talking about are methods, not things. So, let’s say I discover that pregnant women have measurable levels of the hCG hormone, where non-pregnant women do not. Having discovered this, I might try to claim the following:

A method for diagnosing pregnancy in a human female, comprising:

a) testing for the presence of hCG in blood; and

b) diagnosing pregnancy if such hCG is present.

This is not a patent on any particular test – it is a patent on every use of the test that measures for the hormone. Of course, one can measure for hCG without trying to diagnose pregnancy, and that would not be infringing. That said – and here’s the problem for labs – if there is no other purpose for the test, or if the test is ordered for a particular purpose, then one may be liable for administering the test even if one is not doing the last step – the diagnosing.

Among others, there are three primary complaints about these types of claims.

First, some worry that such claims bar thought (the diagnosing). I’m not worried much about such worries. The claim is for more than thought – it is for the ordering of a particular test to diagnose a particular condition (and note that doctors are generally immune from infringement). Also, it should make little difference that the diagnosing step is in the mind – imagine a machine that reads the number and a light goes off (or a stick that shows a plus sign if you are pregnant). Whether you think that these types of methods should be patented or not, the reason to go one way or the other shouldn’t depend on the happenstance of form.

Second, some worry that such claims are simply a patent on the prior art with thought added to the end. For example, if there were pre-existing hCG tests, then one should not be able to patent practicing the prior art with the added “correlative” step of diagnosing a pregnancy. I’m on the fence about this concern. On the one hand, if the solution is obvious, then we shouldn’t allow patents. On the other hand, we have long given patents for new uses of old stuff. This includes new treatments using known medicine – Viagra, for example, was developed to treat blood pressure. We grant such patents because we want inventive activity to find these new uses, and it is not clear why new uses of old tests should be any different. Of course, people may develop better tests, or use existing tests for other purposes.

Third, some worry that this type of claim covers “laws of nature.” This is where the action is. Courts have long said (mostly in repeated dicta rather than actual rulings on disputed patent claims) that laws of nature are not patentable. But what exactly is an unpatented law of nature, and what is a patentable application of a law of nature? Even if we were to accept that laws of nature are unpatentable, the natural law in the above example appears to be the production of hCG by pregnant women. So, if I tried to claim production of hCG by pregnant women, I could not patent that.

Viewed this way, of course laws of nature are not patentable – they aren’t new, they aren’t invented by the patentee, and they lack practical utility because they don’t do anything – they just are. Furthermore, they aren’t processes at all because they are not a series of steps that achieve some end. But when you use the law of nature to provide some new public benefit that is discovered by the patentee, that is an application of a law of nature, and that is patentable. My coauthors and I make a similar argument with respect to the application of abstract ideas in our Stanford Law Review article  “Life After Bilski.” Indeed, there are many, many creative diagnostic/measurement tests dating back to the early 1800’s for measuring one thing by looking at something else that’s “naturally” related. These are not all laws of nature – they are applications of newly discovered laws of nature to a useful end.

Despite the seemingly unassailable logic of the preceding argument, many people (shockingly!) disagree with me. Their argument is that allowing the application of the law of nature would preempt all uses of the law of nature itself, and that is a bad thing. I’m skeptical of this counter-argument for three reasons. One, everything boils down to something natural, and deciding when naturalness ends is too difficult an inquiry. Two, except in very rare cases, a diagnostic test does not bar the natural principle, it just rewards those who figure out why the principle is important. If we are worried about the inability to perform medical tests, then we should handle the problem with compulsory licenses rather than removing the incentive to invest in invention from the equation. Three, there is nothing naturally occurring about the test in the Prometheus case. The drug administered is human-made, and the metabolite measured in the claim does not exist in nature. Thus, testing for the metabolite is something that only happens if you change nature with something unnatural. But it would make little sense for that to be the dividing line on diagnostics – whatever concerns one has about them, they don’t go away if just because they are measuring something that doesn’t happen naturally.

Both those who agree with me and those who don’t are hoping for some clarity from the Prometheus case. Sadly, I think that clarity is not going to come, just like it did not come in Bilski with software and business methods patents. I think clarity is elusive in these cases because clarity is impossible to achieve, as I argue in Everything is Patentable; defining the exception is just too difficult.

But, more than that, clarity won’t come because of the peculiar facts of this case – in short, the patent claim is problematic:

(1) A method of optimizing therapeutic efficacy for treatment of an immune-mediated gastrointestinal disorder, comprising:

(a) administering a drug providing 6-thioguanine to a subject having said immune-mediated gastrointestinal disorder; and

(b) determining the level of 6-thioguanine in said subject having said immune-mediated gastrointestinal disorder,

wherein the level of 6-thioguanine less than about 230 pmol per 8×10^8 red blood cells indicates a need to increase the amount of said drug subsequently administered to said subject and

wherein the level of 6-thioguanine greater than about 400 pmol per 8×10^8 red blood cells indicates a need to decrease the amount of said drug subsequently administered to said subject.

Here are some of the problems:

  1. There is no diagnosing step. It’s a minor point, but an important one. If there is no diagnosing step, then one can infringe the patent by merely performing the test – a test that I believe was already known. It may be a small point to add an element of actually adjusting the dosage, but doing so takes the claim out of the realm of exactly the prior art and into the realm of diagnosing. Bear in mind that Mayo was sued in part because it developed a new test for the metabolite.
  2. This is not really a diagnostic patent. It is not measuring something to determine some condition. Instead, it is a dosage adjustment patent. This, too, should be patentable on general principles, but it seems awfully obvious. If there’s too little, then give more. If there’s too much, then give less. Granted, the patent claims specific numbers, but finding the right range is something that would be obvious to try once you have a test in the first place. Anyone with a thyroid disease knows this – you start with the lowest dosage of synthroid and then go up until the TSH falls within a range.  The patent here is claiming such a range, but finding the range seems obvious- when people get sicker or have side effects, you are outside the range and when people get better, you are inside the range. The real invention is discovering how to measure the metabolite of the drug, but the test for finding the metabolite isn’t a subject of this patent.

These two problems mean that the patent should be rejected outright, and that the patent isn’t really even claiming a natural principle. Instead, it is claiming the use of a pre-existing test and some obvious thought about one might do about the test (but holding people liable even if they don’t do it). Thus, there are many reasons to reject this patent without speaking to the core question – what to do about patenting medical diagnostics. We’ll see how the Court handles it, but my worry is (as it was in Bilski) that a problematic claim will cloud the key issue. A better test would have been a really great, inventive diagnostic test that relies on a natural principle. The Court had such a case in the past (the Lab Corp. case), but procedural failures by the defendant precluded real consideration of the issue.