About the Authors

  • The Authors and Contributors of "Patent Docs" are patent attorneys and agents, many of whom hold doctorates in a diverse array of disciplines.
2018 Juristant Badge - MBHB_165
Juristat #4 Overall Rank

E-mail Newsletter

  • Enter your e-mail address below to receive the "Patent Docs" e-mail newsletter.

Enter your email address:

Delivered by FeedBurner

Contact the Docs

Docs on Twitter


  • "Patent Docs" does not contain any legal advice whatsoever. This weblog is for informational purposes only, and its publication does not create an attorney-client relationship. In addition, nothing on "Patent Docs" constitutes a solicitation for business. This weblog is intended primarily for other attorneys. Moreover, "Patent Docs" is the personal weblog of the Authors; it is not edited by the Authors' employers or clients and, as such, no part of this weblog may be so attributed. All posts on "Patent Docs" should be double-checked for their accuracy and current applicability.
Juristat #8 Overall Rank


« USPTO News: IP Australia Now Competent ISA for PCT Applications Received by USPTO | Main | A Second Look at President-elect Obama's Technology Platform »

November 03, 2008


It is startling to try to even understand the Court's logic in equating Alonso with Rochester. Biotech cases are being unduly punished with the written description requirement, while their mechanical and electrical counterparts get a free-pass.

Dear JT:

I understand the complexity argument, and in context it is a good one. For gene sequences per se, there is a need to actually have the sequence, or to be able to have a representative number, in order for a patentee not to preempt the future by claiming what she hasn't invented.

My issue is that, for making antibodies, once you have the antigen you should have the antibody, absent evidence that the antigen is too weak, or that you need a particularly high affinity/titre for it to be useful, etc. Here, reducing the patentable claims to the disclosed antibody should/could render the claim worthless, because all you need to do is use another antibody to practice the claimed method without infringing.

Now, the case does help because it mentions ways to satisfy even these draconian requirements. There are things the patentee didn't do that could be done to "describe" the antigen, such as Western blots of neurofibrosarcoma cells to identify the protein, or 2-D gel patterns of different neurofibrosarcomas to "identify" which antigen pattern is associated with the antibodies, and of course making more antibodies. But the philosophical point is not to use the written description requirement as a quantitative measure of "how much" but to recognize that both the "what" and the "how much" of a patent disclosure depends on the subject matter. Here, unlike in Rochester, the inventors had provided an exemplar, showing that these cells could be used to raise useful antibodies, in a spec that was enabling. That should be enough.

Thanks for the comment.

"all you need to do is use another antibody to practice the claimed method without infringing"

If it's so easy, then why didn't Alonzo make and characterize additional functional antibodies? Seems trivial.

Unless it isn't.

"the reality is that Dr. Alonso provided a fully characterized plurality of antigens -- the neurofibrosarcoma cell, which despite its heterogeneity is a particularly well-characterized type of cancer cell -- that could be used to reliably produce specific monoclonal antibodies."

I assume Alonso claimed his method of producing specific monoclonal antibodies using that neurofibrosarcoma cell. In that case, he will have blocked his competitors from practicing his surprisingly effective method.

Dear Gwemly:

I wouldn't assume the PTO would grant the method claim.

The point is that you can always impose another requirement on applicants. There seems to me a belief that this doesn't impede innovation. I disagree.

Thanks for the comment.

I wonder if you could help me with the science. As I understand it, Alonso’s method requires production of a specific mAb for each patient (a monoclonal antibody idiotypic to the neurofibrosarcoma of “said” human). Is that right? Also, what exactly was the inventive concept? I take it that the methods of producing a mAb of this general type were well known. Was the basic advance the discovery that neurofibrosarcoma can be treated in this manner?

Dear Norman:

I will admit I haven't read the application, so with that caveat here are my thoughts.

It was well known that you could make antibodies against a foreign cell. I'm not sure that it was known that there would be the kind of discrimination from one human to another. Also, at least in the footnote of the decision, it was remarkable (surprising, unexpected) that it worked so well against at least one patient's tumor.

Since monoclonal antibodies have matured as therapeutic agents, this finding suggests that these kinds of antibodies may be particularly effective against this kind of tumor.

Thanks for the comment.

I don’t think the ‘749 application has been published, but the application from which Alonso claimed priority has issued as Canadian Patent 1341552. (The application is also available as EP0363703.) From what I can understand, the prior art was human-mouse hybridomas and the basic advance was a method of producing human-human hybridomas. I can’t tell what the inventive step was. The method claimed seems to me to be basically the same as that for producing human-mouse hybridomas and the specification does not state any special problem that had to be overcome. So far as I can tell, it was well known in the prior art that problems associated with the therapeutic use of mAbs derived from human-mouse hybridomas arose because of they yielded mouse antibodies, so I can’t image that the notion of using human-human hybridomas was innovative in itself. But I might be judging with the benefit of hindsight.

In any event, let’s leave aside the question of obviousness, which wasn’t in issue. From the specification it seems clear that Gwemly’s suggestion that it would have been trivial to make and characterize additional functional antibodies is wrong. The specification states that one advantage of the method is that it produces aAbs specific to the tumour of the particular patient. Clearly it is not possible to make and characterize mAb specific to all possible cancer patients.

What about the method claims?. The issued Canadian patent and the European application claims “a method for producing human-human hybridomas...” including in a dependent claim a method using neurofibrosarcoma tumour cells, rather than the method of treatment of neurofibrosarcoma at issue in the ‘749. Gwemly suggests that these method of production claims will effectively block competitors from practising the method of treatment. Is that right as a practical matter from an enforcement perspective? I would have thought that there is a practical difference in that the method of treatment claims would allow Alonso to enforce at the point of treatment, while the method of production claims restrict Alonso to going after the producer. That might be a practical problem in some circumstances (foreign production perhaps?) Am I right about this, or do the method of production claims, if issued, give the same practical scope of enforcement as the method of treatment claims?

Keven, you suggested that the PTO might not grant the method of production claims. Why not? I don’t understand the written description doctrine well enough to have much of an opinion on that point.

I should say that I am not suggesting that the written description doctrine is a good thing. I’m just trying to understand the practical implications. And even if the method of production claims give the same effective scope, this does not to my mind justify the written description doctrine. So far as I can understand it, it does strike me that the doctrine is a kind of arbitrary scope restriction. If there is a real difference in enforcement, that’s bad; and if you can get the same protection with a method of production claim, this just makes the restriction on the method of treatment seem even more arbitrary and pointless.


"From the specification it seems clear that Gwemly’s suggestion that it would have been trivial to make and characterize additional functional antibodies is wrong."

For the record, it wasn't me who first suggested that it was trivial to make and characterize additional functional antibodies. That was Dr. Noonan, who wrote that "all you need to do is use another antibody to practice the claimed method without infringing."

I was taking Dr. Noonan's complaint about the "narrow" claim scope to its logical conclusion. Either it's easy to make these antibodies, or it isn't. If it's not at all easy (which Norman seems to think is the case) then Alonzo should be HAPPY with a narrow claim to his very special, useful, and effective antibody.

If it is easy, then we are back to my earlier question: why didn't Alonzo have any additional data?

On re-reading, I agree that you are correct to say that it would be trivial to make and characterize any specific number of additional antibodies, but I think would not be possible to make and characterize additional antibodies that would completely define the functional scope of the claims, because those antibodies are tailored to specific purposes. Alonso is not happy with his narrow claim because that specific antibody is very useful and effective in treating Melanie Brown's neurofibrosarcoma, but it is not useful and effective in treating anyone else's neurofibrosarcoma. That is why the claim is not adequate if it is restricted only to mAbs that are specifically characterized; it would restrict the application to cancer patients actually known to Alonso at the time of the filing.

Maybe I should elaborate on what I gathered from the specification. As I understand it, Alonso's method is to create a human-human hybridoma by priming spleen cells with tumour cells taken from the very patient to be treated (e.g. Melanie Brown); as the claim states the mAb is "idiotypic to the neurofibrosarcoma of _said_ human." The mAb created by the method is then specific to and very effective against that patient's tumour, and it is reinjected into that patient to fight that particular cancer. The specific mAb created in this manner is not necessarily particularly effective against the same type of cancer in a different patient.

Dear Norman and Gwemly:

Sorry for being AWOL from the discussion.

I think Norman is correct in his assessment of the patent disclosure. I think it points to the unreality of the decision. What Dr. Alonzo found was that for this particular type of cancer, neurofibrosarcoma, cells could be taken and used to make human idiotypic antibodies from the patient's own tumor, which could then be used to treat the patient with (admittedly) striking results.

By focusing on bean counting, I think the court missed the significance of the invention. Even should Dr. Alonzo have taken 5 or 10 to 100 patients (disregarding the ethics of such a practice), that would not have satisfied a requirement for written description according to the courts rubrics, because the tumor from the 101st patient would present a different cluster of antigens and hence a different antibody.

Tough to squeeze these facts into the "common structural features" test, and tough even to decide what is a "representative number of species" under these circumstances.

Now, one way to address the issue is to say, "Tough luck" to these kinds of claims, and insofar as there would be a generic claim to an antibody I have some sympathy for the position. But it seems to be a slave to formula to say that the therapeutic method claim is unpatentable just because each patient's tumor will vary unpredictably.

Thanks for the comment.


You and I at least, are on the same page regarding the application of the written description requirement to the therapeutic method claim.

To take one more step, what about the method of production claims? My feeling is that (i) the claims to the method of production of the mAb (as opposed to the method of treatment) would issue - at least they would satisfy the written description requirement; (ii) such claims would provide some degree of protection across the entire scope of the invention, but (iii) as a practical matter the protection would not be as good as that provided by the therapeutic treatment claims. Any thoughts?

I don't want to beat this to death, but I want to confirm that this is a real practical problem and not merely a theoretical problem about the conceptual arbitrariness of the written description requirement.

Dear Norman:

We get back to 35 U.S.C. 103(b), don't we? In the old days, this was the connundrum - an "old" method (here, mAb production) with a "new" result (here, unique idiotypic mAbs effective against a patient's cancer). Sec. 103(b) was the answer (although it isn't used much). Such claims might also have 35 U.S.C. 271(g) implications, so they might be stronger than you would think.

There is a smattering of "angels on the head of a pin" in how the CAFC and PTO apply the written description requirement, but you do illustrate ways for an applicant to get some protection despite this. Of course, the right answer is to apply the requirement logically rather than rigidly in the first place, but that seems to be asking too much.

Thanks for the comment.

Thanks Norman.

Alonzo has provided one example of a useful antibody that shows his method of producing a self-tailored antibody for neurofibrosarcoma is enabled.

Assuming the method is not obvious, Alonzo is entitled to a method claim. The idea that Alonzo should be allowed to reach through and claim **any and every antibody produced by his method** is absurd on its face. I can't believe we're even discussing that possibility in 2008.

Dear Dr. Gwemly:

Not really absurd; the protection should fit the disclosure. What should we do for the following:

I discover a new gene and protein, and I want to claim it. Should I be limited to the nucleic acid sequence of the gene I discovered, or should I be able to recite the nucleotide sequence with reference to the (putative) amino acid sequence? (After all, I didn't sequence thousands of silent variants and the pattern of replacement should be completely unpredictable.)

Now, how about variants of the amino acid sequence? After all, Val and Ile differ by a methylene group. Can I get a claim for my protein wherein any Val has been substituted with Ile? Why or why not?

How much disclosure do I need for other amino acid sequence variants? Is a comparison between/among mammalian species enough?

And we haven't even thought about product by process claims.

The point is that the area is pretty grey where there is a reach-through. Rochester was easy - the claim not only required an inhibitor that was completely unknown, but also that it was a COX-2 selective inhibitor (i.e., it wasn't known whether such an inhibitor even existed).

But here, the discovery is that this particular type of cancer could be treated with idiopathic antibodies. Now, I agree that the applicant could have (and certainly should have) disclosed more and made different/better arguments. My gripe is the mindless recitation of "representative number of species" or "common structural features" when, in a very real way, the "common structural feature" is that the patient has this particular type of cancer.

Reasonable people can differ. Thanks for the comment.

Dr. Gwemly,

I'm afraid I don't see why the "reach through" claim is absurd. Suppose Alonso had discovered a novel method for making a single novel single compound that cured all neurofibrosarcomas. In that case I think it is clear that he could have a valid patent for the compound as well as the method. I don't see why reaching through to a patent for the compound should be any more objectionable simply because the compounds produced by the method are tailored to the individual. In either case the scope of the contribution is the same - he has enabled a cure for all individuals - and it seems to me that the scope of his patent should be commensurate with the scope of his contribution.

"Suppose Alonso had discovered a novel method for making a single novel single compound that cured all neurofibrosarcomas. In that case I think it is clear that he could have a valid patent for the compound as well as the method."

Uh ... isn't that essentially what happened here? Except that Alonzo's novel method may not be non-obvious.

"I don't see why reaching through to a patent for the compound should be any more objectionable simply because the compounds produced by the method are tailored to the individual."

Any more objectionable than what? The law requires that the claimed composition be conceived in a fixed and permanent matter in way that distinguishes it from the prior art and the inventor must show that he was in possession of a reasonable number of those new, non-obvious compositions by describing them with sufficient particularly. As a matter of objective fact, Alonzo does not have that information. He does not and can not know what the "special" characteristics of the functioning antibodies are. All he knows is how one would reasonably attempt to make such an antibody.

"he has enabled a cure for all individuals "

If so, then he is entitled to a method of treatment claim. If he is not happy with that, then he doesn't have to file a patent. He can just be happy with his Nobel prize and the glory associated with his life-saving achievement.

Kevin: "I agree that the applicant could have (and certainly should have) disclosed more and made different/better arguments. My gripe is the mindless recitation of "representative number of species" or "common structural features" when, in a very real way, the "common structural feature" is that the patient has this particular type of cancer."

Even if I grant you that "having a particular type of cancer" is a "common structural feature", that's light years away from the sort of structural information necessary to describe the essential structural features of an antibody that will have the functionality of the single mAb that Alonzo has reduced to practice.

The reason composition claims are desired is because they preclude individuals from making the claimed compositions using any method, not just the method described in the patent. You seem to think that Alonzo should be given, effectively, an old-school product-by-process claim that would cover any mAb that functions in the way that Alonzo's mAb functions, whether made by Alonzo's process or not.

Reasonable people may disagree, but I can tell you now with 100% certainty that you will not live long enough to see such a claim enforced in this country and upheld by the CAFC. We know now that product-by-process claims are actually poorly written method claims.

Dear Gwemly:

"If so, then he is entitled to a method of treatment claim."

Agreed. But the problem is that the CAFC said he is NOT entitled to a method of treatment claim UNLESS the method is expressly limited to the disclosed antibody. Since the point of the invention is that each antibody is idiotypic for not only the particular patient but for that patient's tumor, limiting the patentable claim in this way is equivalent to no claim at all.

Now, a claim to a pharmaceutical composition comprising an idiotypic monoclonal antibody immmunologically reactive with a neurofibrosarcoma is the kind of "reach through" claim you seem to find objectionable, and there you are probably right on these facts. But the CAFC uses Rochester to say that they have to treat the method of treatment claim the same way as they treat my hypothetical composition claim, because if they don't everyone will merely cast unpatentable reach through composition claims as method claims.

That made sense in Rochester on its facts; here, the distinction is that Dr. Alonzo's specification was enabling (which it was not in Rochester) and Dr. Alonzo had shown possession of an exemplar of the mAbs useful in practicing the claimed methods (also missing from Rochester).

So the court had ample bases for making a distinction. My objection is that instead of telling us why they did not distinguish, they merely applied the Fiers/Lilly/Enzo/Rochester/Carnegie line of cases to reject on an almost per se basis. That's what I disagree with.

Thanks for the comment.

Dear Gwemly:

I think product-by-process claims have their place, but that is another example of hobgoblin consistency getting in the way of clear thinking. There is nothing inconsistent (on a policy basis) with saying that the product of a product-by-process claim can be/should be anticipated by that product being known in the prior art produced by any other process, and at the same time, limiting the product of a product-by-process claim to those made by the "same" process in litigation.

The reason I think this is that the product-by-process claim is a product claim, just one where (historically) it was impossible to identify the product with the type of clarity usually found in chemical compound claims. (After all, you never see a product-by-process claim for a bicycle.)

BUT once such a claim issues, the capacity to produce a specific product falling within the scope by a different process should not be an infringement, since a good portion of what the patentee discloses in fulfilling her part in the patent bargain is an enabling process for making the product. This is a result that promotes "designing around," wherein the impetus exists to do the analytical chemistry (or here, the analytical immunology) to identify just the kind of common structural features that would lead the skilled artisan to be able to produce non-infringing composition claims.

But I still think Dr. Alonzo was entitled to the method of treatment claims, for all the reasons upthread.

Thanks for the comment.

I just wanted to circle back and say that, while we disagree on the composition claim, I think I do agree that a method of treatment could have been patentable here had the patentee provided just a bit more data, either at the time of filing or afterwards, to show that the method was generally applicable (i.e., to enable the broad scope of the claims). I disagree with the use of the written description test in this case. It's an enablement issue. It's a somewhat close case of enablement, but I think it is reasonable to request at least one more working example from the inventor given the "individual" nature of the treatment method.

The comments to this entry are closed.

August 2020

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31