Applying Bilski to Metabolite’s Diagnosis Claim

Metabolite’s broadest method claim has two steps: First, assay a body fluid to determine its homocysteine level. Second, correlate the homocysteine level to a vitamin deficiency. It reads as follows:

13. A method for detecting a deficiency of cobalamin or folate in warm-blooded animals comprising the steps of: assaying a body fluid for an elevated level of total homocysteine; and correlating an elevated level of total homocysteine in said body fluid with a deficiency of cobalamin or folate.

In 2006, the Supreme Court almost took the time to reject this claim under Section 101, but decided that the issue had not been properly raised on appeal. Dissenting from the last minute denial of certiorari were Justices Breyer, Stevens and Souter. In a dissenting opinion drafted by Justice Breyer, the three would have held the invention as claimed an unpatentable “law of nature.” The same anti-preemption theme of Bilski also runs through the Breyer opinion. The courts are concerned with the potential (but unidentified) problems of having a patent that

There can be little doubt that the correlation between homocysteine and vitamin deficiency set forth in claim 13 is a “natural phenomenon.” . . . .

Claim 13 . . . tells the user to use any test at all. Indeed, to use virtually any natural phenomenon for virtually any useful purpose could well involve the use of empirical information obtained through an unpatented means that might have involved transforming matter.

I thought it would be interesting to attempt to apply the Bilski decision to Claim 13 to judge its patentability.


Bilski offers two avenues for showing that a claimed process is patentable. First, the process will be patentable if it is tied to a particular machine or apparatus. Alternatively, the process will be patentable if it “transforms a particular article into a different state or thing.” According to the court, “particular article” does not necessarily need to be physical, but it must at least be “representative of physical objects or substances.”

Machine: Metabolite’s Claim 13 easily fails the first prong of the machine-transformation test because the claimed method is not tied to any particular machine.

Transformation: The “correlating” step could be performed in a human mind and consequently cannot serve as the basis for finding patentable subject matter. As the court noted in Bilski – a process step that “may be performed entirely in the human mind is obviously not tied to any machine and does not transform any article into a different state or thing.”

The “assay” step could follow the same analysis – if it is possible to “assay a body fluid for an elevated level of total homocysteine” only using the human mind then no there is no credible transformation. The mental step potential may be true if, for example, the body fluid’s homocysteine level is visually apparent. [Interesting timing issues here, at the time of the invention, the assay involved extensive lab work, but later-invented technology has dramatically changed the process.]

Assuming that “assaying a body fluid for an elevated level of total homocysteine” cannot be done in the mind, we must then ask whether the claim requires a transformation of a particular physical object or substance or at least transformation of a particular article that is “representative of physical objects or substances.” Perhaps the best physically related transformation involves transforming the “body fluid” into an indication of elevated homocysteine level and then into an indication of vitamin deficiency. Under Abele, both homocysteine level and vitamin deficiency do represent physical substances – consequently their ‘creation’ may be “sufficient to render that … process patent-eligible.” [Of course, this analysis may apply a loose definition of 'transformation.']


29 thoughts on “Applying Bilski to Metabolite’s Diagnosis Claim

  1. Eye: “Mooney’s genius communication step is so much insignificant post solution activity so as to be, using words that Mooney understands all too well, doo doo.”

    I am still waiting to hear why a communication step is so “insignificant”. Evidently, it’s significant enough to cause people to worry that requiring the step will destroy the value of the patent and cause untold human misery. And I think we’d not have heard of LabCorp. had such a step been put in claim 13 from the beginning.

    So why is it insignificant?

  2. “I think Malcom’s claim has the best chance of surviving Bilski”

    You must be kidding me. Mooney’s genius communication step is so much insignificant post solution activity so as to be, using words that Mooney understands all too well, doo doo.

    I sincerely hope that everyone here continues to listen carefully to Mooney so that when everyone who does eventually gets sued for malpractice my base of potential new clients will, to my great delight, increase dramatically.

    The key would be to break the assay into two or more transformative steps, but that is all I’m saying. I’m done giving valuable lessons for free here on Patently-O.

    Please continue to get all your free patent law education and insight from m0r0ns like Mooney (who, incidentally is not a patent attorney, but sweeps floors in the metro station). And best of luck to you.

    With the recent Patently-O post regarding fees in mind. Remember, you get what you pay for.

    There is such a huge gulf between poseurs like Mooney and people who really know what they’re doing, like, for example, John Darling.

  3. I wanted to post simply to echo Phil Marcus’ comments.

    The “natural law” of special relativity (which everybody loves to use as an example of patent-ineligible subject matter) is qualitatively VERY different from the “natural phenomenon” of elevated homocysteine levels being associated with folate deficiency.


  4. Since the claim does not literally and specifically tie the process steps to a machine, there is no clear basis to disclaim processes that are solely mental ones. Since claims have to be construed to their broadest possible limits, the inherent point of Justice Breyer’s comment, the claim fails under Bilski.

  5. Here are my two cents:

    Any claim requiring BOTH a physical assaying step and either a mental correlation step or an arguably physical communication/recording/prescribing/treating step is way too easy to get around (what single party directly infringes?) I assume this is what Steve Sereboff meant by “BMC v Paymentech issues.”

    Virtually all assaying steps will involve a transformation of an article (amplification of DNA, mass spec on a compound, etc.). However, as the quote from Breyer’s dissent above suggests and Bilski slams home, such data-gathering will be insignificant extra-solution activity (assuming the assay is not novel). Similar result (though different analysis) to Comiskey’s pronouncement that addition of a conventional machine or transformation might get you out of 101 but doom you under 103.

    That some part of the claim may be practiced in the mind is not dispositive. That the crux of the invention (i.e., the only novel aspect) may be practiced entirely in the mind may be dispositive. Recall that Benson’s claims failed despite the addition of a computer because the claims could only reasonably be practiced using a computer. Does this mean a multivariate correlation is pinched (i.e., might have been patentable because it more clearly represents a transformation of data representing some physical element, but fails under Benson because it can only reasonably be practiced by computer)? Just something we have been grappling with.

    Ultimately, I think Malcolm’s claim (assay plus communicating/recording/treating) has the best CHANCE of surviving Bilski. However, its preemptive footprint is negligibly smaller (if at all) than the Metabolite claim (policy problem). Also, as noted above, no single party need ever infringe the claim. Then you can throw in the uncertainty regarding 271(f) and foreign outsourcing of the assaying step (paired with the fact information is not a “product” under 271(g) and Bayer v Housey) and you start to get the feeling the CAFC (except maybe Rader or Newman) didn’t fully realize the destructive power of the Bilski decision (or at least HOPE they didn’t).

  6. As the only patent attorney at my company, I miss discussions like this. Thanks for the dissection of this issue.

    The problem I am having here is that correlation itself is also a known method. It is also one that is susceptible of being carried out in one’s mind. In my opinion the only patentability here would be found in a novel assay or treatment based on the correlation. The problems for enforcement of a claim like this have already been pointed out.

    I also would like to point out that there are many, many motivations for a company or researcher to develop, publish, and/or implement assay/treatment regimes such as the one described, even if it is not patentable. The fact that people might die if the regime is not implemented or published is one of them. I understand the dis/incentives companies are working under and I feel that the incentive to practice something like this simply because it can be practiced without fear of infringement (presuming that there is some efficacious effect) outweigh the disincentives to practice based on lack of IP protection. This is not always the case but I feel that the idea that it is only IP protection that gives anyone the impetus to be creative or to develop something more fully is incorrect. It is a factor, but there are other reasons too.


  7. In my view, the problem with both Bilski and the SCOTUS almost decision two years ago is that they are discussing section 101 as if it were a Congressional exercise in metaphysics. This is very unfortunate for the nation’s economy, and a misreading of the framers’ words. The Constitutional provision on patents and copyrights was clearly an economic matter, not a metaphysical or philosophic issue they could have discussed over a whiskey and water privately. And presumably Congress has extended patents and copyrights to the Constitutional limits.

    Discovering that a concentration of a molecule in the bloodstream is an assay for a disease is nowhere in the neighborhood of discovering, for example, the law of conservation of momentum. The latter is a law of nature. The former is a creative act deserving of protection, in the interest not of the inventor but of the nation.

    Congress needs to rewrite section 101 to straighten out both these courts, getting them out of their ivory towers.

  8. Sorry–I see that “divided infringement” case was BMC v Paymentech, as noted earlier by Steve Sereboff. Not sure how you could get around that.

  9. Of course the claim is patentable subject matter. Is there a transformation step? Yes there is — the assay is a transformation (it doesn’t matter that the assay is old). Is there a mental step? Yes, but a single mental step doesn’t make a claim unpatentable.

    The claim may be unpatentable but not under 101. Also, the claim may be patentable but not particularly enforceable. If a lab is doing the admittedly old assay and isn’t telling the doc what to do with the information, then the patent holder will have to go after the docs. Ain’t gonna happen.

    To those who go on about the claim being “unpatentable as a law of nature”, that’s not correct. “Correlating” homocysteine content with a vitamin deficiency is not a law of nature because it requires at the least a mental step. If you take the position that correlating two things are a “law of nature” then what are you left with? Electrons moving around on a chip (or whatever they do) is a law of nature.

  10. James Daily–I like this approach, although I agree with others that it is not quite there yet. Maybe a “whereby” clause in step (2)–like whereby a treatment is selected to ameliorate the effects of said disease?? Also one would need to keep in mind that if different steps were performed by different entities (say, the doctor and a testing lab), then proving infringement might not be possible–I forget which recent case held this (maybe a computer-related case?) but that could also be a problem.

  11. While one can apply Bilski, I agree with the comments that the claim is unpatentable as a law of nature. I.e., merely “correlating” substances and properties can’t be patentable, and that is essentially what this claim does.

  12. I don’t know if Bilski is needed to limit “unnecessary” software patents, but it does lead to what I consider an undesirable result for medical diagnosis.

    James Daily at 6:17 PM
    Some diagnoses might not require treatment but only a change in lifestyle, for instance avoiding eating gluten or nuts.

    But importantly, I was trying to say that you want to keep your claims away from actions performed by medical personnel and towards actions performed by manufacturers and suppliers.

    Malcom 6:27 PM
    Surprisingly, I was slightly involved in three cases which were diagnoses based on an observation of a correlation.

  13. Erez “As Piled Higher and Deeper and I wrote in the November 4 “Applying Bilski..” thread, the only apparent effect is that life saving diagnoses won’t be developed and, as a result, people will die.”

    Oh, baloney.

  14. ???? “If the mental step may be difficult or even impossible to prove relevant to patentability?”

    If I understand the question (a big if), the answer is yes, insofar as the prohibition against mental steps is in part a recognition of this problem (as well as, I suppose, privacy concerns).

  15. In my example, we may presume that the mental step is complicated enough that it would leave certain identifiable traces. For instance, noting certain information in the patient’s chart, requiring a computer program to perform the calculations, etc. Or, perhaps the mental step is the only known way to diagnosis Disease X from the physical test (as a contrived example, diagnosing brain cancer from the results of a foot x-ray). This would provide circumstantial but persuasive evidence that the physician performed the claimed mental process.

    In any event, if it is difficult to prove the mental step for purposes of proving infringement, then the patent is weak and not very valuable. But if that is the case, then it is unlikely that anyone would pursue the patent in the first place. I don’t think it is reasonable to eliminate a whole class of diagnostic patents simply because some of them might not be very valuable. The market will keep many of those out.

    Ultimately what I’m getting at is that it seems to me that there should still be a way, post-Bilski, to claim new uses for the data produced by existing tests. I think it would be an unreasonable to result to limit diagnostic patents only to those that claim a novel physical test. Medicine has a limited toolbox, and we should encourage the discovery of new uses for existing tools.

    As an aside to Erez Gur: You suggest that appending a treatment step is inadequate because Big Pharma has all of the treatments locked up. This is not necessarily so: consider the hundreds of drugs and surgical treatments that are out of patent. Furthermore, treatment of some kind necessarily follows diagnosis; what use is a diagnosis without treatment? I don’t think adding the treatment step substantially limits the scope of the claim as long as the treatment step is written broadly (e.g. “administering an effective treatment for Disease X to the patient” or somesuch).

  16. Bilski seemingly kills claims such as:
    A diagnosis based on using a well-known assay to ascertain a concentration of a metabolite in a blood sample, where the novelty is interpreting a specific concentration as indicative of a disease.

    It is not clear why people like Captain Obviousness are so happy with this result.

    As Piled Higher and Deeper and I wrote in the November 4 “Applying Bilski..” thread, the only apparent effect is that life saving diagnoses won’t be developed and, as a result, people will die.

    James Daily’s solution is also inadequate, whether the patient is transformed or not. The only person you could sue is medical personnel who practice claim 3. It likely that step 3 is patented by someone (e.g. Big Pharma), in which case you can’t get investors because you can’t practice your own invention. Of course, such a claim would be a great way for Pharma to evergreen their compositions.

  17. If the mental step may be difficult or even impossible to prove relevant to patentability? Or is it only relevant to infringement?

  18. JD “A method of diagnosing Disease X in a patient comprising the steps of:

    1. Gathering data by performing a specifically claimed, undeniably physical lab test that is known in the art;

    2. Using a disclosed novel mental process to determine from the results of the test that the patient has Disease X;

    3. And administering a specifically claimed, undeniably physical treatment that is known in the art to the patient.”

    I think this gets you out of Bilski as sufficiently transformative but if the assay and the treatment are both old, then I believe that your mental process will not save you. Consider: how do you prove the mental step?

  19. JJL “Even if the correlation was well known and it was the assay that was new, the claim as drafted doesn’t claim a particular assay.”

    Well, let’s say hypothetically that “homocysteine” was a completely unknown molecule prior to applicant’s discovery. Nobody had seen it before or, if it was ever “detected” it was considered to be an artefact and not worth measuring.

    But now that the skilled artisan (including applicant) knows what it is, it can be measured by all kinds of assays, including patentable assays that haven’t even been invented it. This is true of most newly discovered molecules that are found naturally.

    Assuming that the molecule can’t be measured without an assay of some sort (i.e., it can’t be measured by eyeball), are you saying that the claim as written would still not fly under *Bilski*? I’m not sure I agree.

    And if you’re correct, is there no way to fix the claim to meet Bilski’s standard other than reciting a specific assay? Or do you think that adding a communicating step suffice?

  20. I have a question for those who find the claim obvious because the assay is insignificant extra-solution activity and the method as a whole is insufficiently transformative. Could this defect be cured by appending a step describing administering a treatment for the diagnosed condition? A generic example follows:

    A method of diagnosing Disease X in a patient comprising the steps of:

    1. Gathering data by performing a specifically claimed, undeniably physical lab test that is known in the art;

    2. Using a disclosed novel mental process to determine from the results of the test that the patient has Disease X;

    3. And administering a specifically claimed, undeniably physical treatment that is known in the art to the patient.

    I would argue that this outline represents a transformation of a particular article (the patient) from a sick patient into a well patient, which is a concrete change of physical state. Is that a reasonable conclusion? Would such a patent be valid after Bilski?

  21. Ineligible because the only statutory step is a mere data gathering step. Gathering input data for the mental correlation does not place any limit on the scope of the claim. If you want to perform a mental correlation, obviously you need an input for the correlation.

    Good riddance to this garbage, and may Bilski be extended as much as possible.

  22. I mostly agree with the other posters except that I think under Bilski it doesn’t matter what the “novel” step is. Even if the correlation was well known and it was the assay that was new, the claim as drafted doesn’t claim a particular assay. It claims a correlation based on measurements taken using any assay at all. That’s why the assay is insignificant extra-solution data gathering and not a transformative step.

  23. I say it is not patentable under Bilski. Here is how I see the analysis.

    Clearly it is not tied to a particular machine.

    I understand the assaying step to be a physical, but unspecified, test of the body fluid. Therefore, I don’t think it is a mental step. However, it would merely be insignificant extra-soloution activity. It’s just data gathering, which would not be sufficient to make the mental step of corelating–which is where the novelty resides–patentable.

  24. Bilski held that the transformation or apparatus that satisfies the § 101 test (machine or transformation) must also “impose meaningful limits on the claim’s scope to impart patent-eligibility.” I would assume that a data gathering step (such as assaying) would fail to qualify as a meaningful limit. Thus, I don’t think that you have a machine or transformation (I assume that “correlating” couldn’t suffice as a machine or transformation).

    I also think that, since Bilski expressly stated in a footnote that an examiner doesn’t have to look to 101 first, the claim would be rejected under 103 before it was rejected under 101.

  25. One observation from the discussion in the “applying Bilski to biotech” thread is that there are at least two distinct scenarios that should not be confused with one another. They are both interesting but the issues are different.

    Situation One is: the Metabolite situation where the claimed assay is old (or left undefined) but the “correlation” between the assay result and the disease is new, will recitation of a step of “recording” a diagnosis based on the assay be sufficient transformative and.or non-pre-empting to get around Metabolite/Bilski? Or is “communication” of the diagnosis sufficiently transformative and/or non-prempting?

    Situation Two is where the assay itself is new (maybe it recites a new antibody or a new method of preparing serum samples). The question is: are there classes of claims to new assays that are affected by Bilski, such that additional steps may be required?

    I agree with Dennis that there is a problem if the new assay can be practiced mentally. But I don’t think that problem is unique to the post-Bilski world. It’s a Situation Two problem, but probably the easiest one to deal with: add a non-mental step of communicating the result.

    “Under Abele, both homocysteine level and vitamin deficiency do represent physical substances – consequently their ‘creation’ may be “sufficient to render that … process patent-eligible.”

    I don’t think so. The homocysteine levels are inherent to the patient. They are not “created” by the patentee or the claim. And the whole point/problem of Metabolite is that once the information about the correlation is in the public sphere (i.e., once the application is published), then “homocysteine levels” and “vitamin deficiency” are inherently linked … related, if not equivalent. The only transformation is the transformation of an idea (“homocysteine level”-> “vitamin deficiency”) in the mind of the person observing the assay results.

    This is why I favor the communication step … at least until a court shoots it down. ;)

  26. Thought the problem with claim 13 was the same issue raised in Comiskey -

    combination of 101-ineligible subject matter (“correlating” mental step) with 103-obvious subject matter (generic assay of which at least one species was known in the art).

Comments are closed.