Examiner Rejects Compound Claim as Anticipated by Error in Prior Art
I recently worked on a case where the examiner rejected a claimed compound as anticipated by an incorrect structural formula. The case made for an interesting case of “anticipation” under 35 USC 102.
Conventionally, an examiner rejects a compound as anticipated upon finding that compound in a structural database. But, this time, the structure was clearly wrong. The author assigned the wrong structure to a molecule that he made. Based on common sense, rejecting the compound for lack of novelty seemed improper because the rejection was based solely on an error cataloging a compound by the wrong structure. The claimed compound was—in fact—new.
During prosecution, all parties knew the right result. The examiner wanted to allow the claim. But the examiner could not get around finding an exact structural match. The applicant, knowing he was the first to make the compound, was understandably frustrated by the Examiner’s insistence that his invention was “not new.”
Error in Prior Art’s Structural Formula
Here are a few relevant facts:
- The claim at issue was to a chemical compound, defined by a generic structural formula (“the inventor’s claimed genus”).
- The prior art disclosed a specific structural formula falling within the inventor’s claimed genus.
- The prior art structural formula was wrong. The author of that art made a mistake when assigning a structural formula to a compound that he made.
Structural Match Found by Examiner
When searching for the claimed compound by structural formula, the examiner finds a structure that clearly falls within the claimed genus. The examiner issues a rejection for anticipation under 35 USC 102 (a). A disclosure of a chemical species anticipates a genus comprising it. “A generic claim cannot be allowed to an applicant if the prior art discloses a species falling within the claimed genus.” The species in that case will anticipate the genus. In re Slayter, 276 F.2d 408, 411, 125 USPQ 345, 347 (CCPA 1960); In re Gosteli, 872 F.2d 1008, 10 USPQ2d 1614 (Fed. Cir. 1989).
Section 102 (a) provides that “A person shall be entitled to a patent unless the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.” From the examiner’s perspective, the scientific literature showed a clearly anticipatory compound.
The claimed compound is new, despite the structural match.
Upon receiving the Office Action, the inventor looks at the prior art and immediately sees the mistake. The inventor is the world’s expert in the class of molecules at issue. He provides an affidavit explaining the mistake. He explains that it would be impossible for the prior art’s disclosure to provide his claimed compounds.
When arguing fails to remove the rejection, the inventor experimentally confirms the mistake by reproducing the prior art’s procedure for making the compound. This evidence also does not remove the rejection. The examiner reasons that an ordinary artisan would have recognized how to make the claimed compounds based on (a) the prior art’s disclosure and (b) ordinary skill. (Notably, this conversation is directed to whether the prior art was enabling).
Arguing Novelty over a Structural Error
Below, I describe how an applicant can work with the case law to remove the rejection. The theory relies on two firmly established principles in patent law. First, a “compound” is not limited to its structural formula but includes all aspects of the compound. Second, anticipation requires prior art disclosing exactly the same invention found in the claim at issue.
First, remember that the term compound is a legally significant term. It means the compound and all of it’s properties. “From the standpoint of patent law, a compound and all of its properties are inseparable; they are one and the same thing.” In re Papesch, 315 F.2d 381, 391 (CCPA 1963).
Also, anticipation means that the prior art compound is exactly the same as the claimed compound. “[A]nticipation under § 102 can be found only when the reference discloses exactly what is claimed and that where there are differences between the reference disclosure and the claim, the rejection must be based on § 103 which takes differences into account.” Titanium Metals Corp. v. Banner, 778 F.2d 775, 227 USPQ 773 (Fed. Cir. 1985).
Comparing Structural Formulas
Comparing any two compounds could lead to four results because a compound includes (1) structure and (2) properties:
- Two compounds that are exactly the same (same structure and properties)
- Two compounds differing in structure only
- Two compounds differing in structure and properties
- Two compounds differing in properties only
Anticipation requires option number #1. Each of the other three scenarios include some differences, which should move the analysis into the framework of obviousness under 35 USC 103.
A case where the prior art only discloses the structural formula (by mistake) fails to provide a compound having the same properties as the claimed compound. Accordingly, that case falls into scenario #4, above. Because the properties are different, the compounds are different. Any difference means no anticipation. All differences must be considered under 103.
I would welcome other thoughts on this sort of rejection. Please feel free to add your thoughts by using the comments section below.
Picture of the 23andMe product, from their webpage
23andMe, Inc.’s Patented “Designer Baby System.”
Yesterday Fox News ran the headline “Genomics firm 23andMe patents ‘designer baby’ system, promises not to use it.” Many other media outlets have adopted the phrase “designer babies,” offering all sort of moral and ethical commentary on whether humankind should be playing God and tinkering with Genomics. I will avoid all of these issues in this article because I don’t find them interesting. This article will focus on 23andMe’s patent. What did 23andMe patent? What didn’t they patent?
23andMe’s U.S. Patent No. 8,543,339
On September 24, 2013, 23andmMe, Inc. was awarded a patent on their “Gamete Donor Selection on Genetic Calculations” application. Their application (U.S. Patent App. No. 12/592,950 claiming priority to U.S. Provisional Patent App. No. 61/201,101, filed Dec. 5, 2008) issued as U.S. Patent No. 8,543,339 (“the ‘339 Patent”). With the award of the patent, various news organizations have become interesting in 23andMe and it’s patents. Below, I offer my thoughts as a chemical patent expert without any interest in the ethical and moral implications of this technology.
23andMe’s Patented Property
The claims in a patent define the property granted in the patent. All of the articles about 23andMe’s patent discuss what the patent describes or “lays out” in the specification. However, only the claims define the metes and bounds of 23andMe’s property. The claims are found at the end of the patent document, just after the heading “What is claimed is:”
U.S. Patent No. 8,543,339 has 28 Claims
The ‘339 Patent has 28 claims. These claims define three different inventions at various levels of generality.
- Claim 1 defines “a system for gamete donor selection….” Claims 2-10 and 22 define variations of that general invention by dependent claims.
- Claim 11 defines “a method for gamete donor selection….” Claims 12-20, and 23 define variations on that general invention by using dependent claims.
- Claim 21 defines “a non-transitory computer program product for gamete donor selection….” Claims 24-28 define variations on that general invention by using dependent claims.
In short, the claims in 23andMe’s patent define systems, methods, and computer programs for gamete donor selection. This should be kept firmly in mind before the company is accused of “owning” babies, genetic material, or the like. It’s also interesting to note that the product they sell isn’t really a gamete donor selection product.
All Claims in 23andMe’s ‘339 Patent are Limited to Identifying a Preferred Donor from a Plurality of Donors
This is where it gets interesting. All of the claims in 23andMe’s ‘339 patent are limited to systems, methods, and programs that include identifying a preferred donor among a plurality of donors based on 23andMe’s statistical analysis. In short, without “identifying a preferred donor,” a system, method, or program would apparently not infringe on the claims in the patent.
On October 1st, 2013, 23andMe published an article stating the company’s position on the patented technology. In short, the company is completely dedicated to “enjoyable”genetic exploration,” and offering people with an “engaging way” to “better understand” genetics by introducing people to their own DNA. The company disavows any intentions to actually pursue the invention described and claimed in the ‘339 patent.
23andMe Patent Covers the Controversial Subject Matter—But the company doesn’t practice that invention.
Notably, the service provided by 23andMe does not include “identifying a preferred donor.” Accordingly the product that 23andMe sells on its website is not covered by the ‘339 patent. I find this fascinating: the company sells a really cool product that is not patented and explicitly avoids monetizing the intellectual property that it has patented. This raises two questions for me: (1) Is 23andMe pursing patent coverage on the product that it is presently marketing and selling? and (2) what does it intend to do with the potentially valuable intellectual property that it has recently received?
First, the home testing product could be very valuable. It does offer a fun and engaging way to learn about your own DNA. Why not pursue claims and draft an application covering that product? It has already demonstrated tremendous value in the marketplace. Plus, all of this recent press should lead to a massive increase in demand.
Second, what is going to happen with the patented technology? Despite issuing a press release stating that the company would never enter the designer baby market itself, the company did not say anything about licensing or selling the technology to another entity. I suspect that the market for designing a healthy baby with desired physical attributes could be quite lucrative. 23andMe says that they “want to be very clear about our technology and intentions.” In particular, aside from spending 6 years drafting and prosecuting a US Patent application, they have “never pursued” the designer baby market, “nor do [they] have any plans to do so….”
Do those statements include selling or licensing the technology to other companies?
Does U.S. Patent No. 6,630,507 Give the Government Ownership of Marijuana?
No. U.S. Patent No. 6,630,507 does not give the government ownership of the marijuana plant. The claims in the patent define methods of using certain non-psychoactive cannabinoid compounds for treating a list of diseases caused by oxidative stress.
U.S. Patent No. 6,630,507 is limited to methods of using a certain class of isolated cannabinoid compounds—it does NOT claim the plant (or any compounds in the plant) per se.
Marijuana Technology and Marijuana Patents
I recently worked on a project having to do with a new marijuana technology. During the project, I became interested in whether entrepreneurs had begun mining the marijuana patent landscape. I asked Google. When I searched for “marijuana patents,” the most popular result had to do with the U.S. Government (the “government”) owning marijuana by virtue of a U.S. Patent No. 6,630,507. This notion appears to have many activist groups upset. Those groups believe that the government is limiting their use of marijuana through patent enforcement. Is this possible?
No. The government does not “own marijuana” for several reasons.
The marijuana plant is not patentable because it has been in the public domain for centuries.
Products of Nature are Not Patentable
First, marijuana is a natural product. It has been around forever. But under US patent law, only new things can be patented. Since marijuana is not new, it cannot be patented. Naturally occurring substances like marijuana are in the public domain. I recently looked into a similar uprising about fennel, which is also in the public domain. Things in the public domain belong to the public without any patent-related royalties or exclusivity. From the standpoint of patent law, we all own marijuana communally.
U.S. Patent No. 6,630,507 Only Claims Methods of Use
Second, every claim in U.S. Patent No. 6,630,507 defines a method of use. Not a compound. Not a composition of matter. The government does not own the patent rights to marijuana per se. It owns the rights to methods of using certain cannabinoids for treating oxidative stress. The government could prevent another entity for methodically using certain cannabinoids for the purpose of treating a disease caused by oxidative stress.
Although marijuana (the plant) is not patentable, various marijuana technologies can be patented. In the case of U.S. Patent No. 6,630,507, the inventor patented a method of using certain (isolated) cannabinoids for a specific medical treatment.
The claimed methods are limited to treating certain diseases
According to U.S. Patent No. 6,630,507, the cannabinoids used in the claimed method are found to have particular application as neuroprotectants, for example in limiting neurological damage following ischemic insults, such as stroke and trauma, or in the treatment of neurodegenerative diseases, such as Alzheimer’s disease, Parkinson’s disease and HIV dementia.
Notably, about 1 hour ago, the Huffington Post ran a story about an “Ohio Mother Moving to Colorado to Treat Daughter’s Epilepsy With Cannabidiol Oil.” That particular use could arguably fall within the claims of the patent. The isolated oil seems to be within the scope of the nonpsychoactive compounds identified in the patent. Plus, treating epilepsy seems to be within the scope of the methods of treating ” neurological damage following ischemic insults, such as stroke and trauma, or in the treatment of neurodegenerative diseases.”
On the one hand, folks will probably get bent out of shape if the government changes the mother a royalty on using the isolated oil to treat her daugter’s epilepsy. On the other hand, the inventors of that patent made that treatment an option and arguably deserve a royalty for their contribution.
The U.S. Patent System does not offer a route to “own” marijuana because that plant is in the public domain
The Claimed Methods are Limited to Administering NonPsychoactive Cannabinoids.
The claimed methods use a class of cannabinoid compounds that avoid the “high” traditionally experienced by marijuana users. In particular, “Nonpsychoactive cannabinoids, such as cannabidiol, are particularly advantageous to use because they avoid toxicity that is encountered with psychoactive cannabinoids at high doses useful in the method of the present invention.”
The patent further distinguishes the cannabiniods used in its methods from other cannabinoids that produce “undesired psychotropic side effects.” The invention appears limited to methods of treating oxidative stress with a subset of such drugs are “substantially free of psychoactive or psychotoxic effects.” As a back of the envelope test, a marijuana user should could determine that she is not infringing on the 6,630,507 patent any time she experiences “psychoactive” properties.
Going back to the example of the Ohio mom from the Huffington Post article: from the standpoint of patent law the government would seem to have a better case for preventing the girl from using isolated nonpsychoactive cannabidiol than using naturally occurring marijuana for the same purpose. Part of the government’s invention seems to be separating the therapeutic (neuroprotective) properties from the unwanted psychoactive properties found in the natural product.
U.S. Patent No. 6,630,507 has 26 method claims.
Claim 1 is probably the broadest claim. It recites:
A method of treating diseases caused by oxidative stress, comprising administering a therapeutically effective amount of a cannabinoid that has substantially no binding to the NMDA receptor to a subject who has a disease caused by oxidative stress.
This is a method of use claim. The government does not own marijuana. Rather, the government owns the rights to a method of using a particular group of cannabinoid molecules to treat “oxidative stress.” Oxidative stress is defined in the specification. The group of cannabinoid molecules is limited to those having “substantially no binding to the NMDA receptor.”
Stop Blaming Patent Law for Marijuana Restrictions
The fact that the government does not “own” marijuana by virtue of s U.S. Patent No. 6,630,507 does not preclude the government from prohibiting people from using marijuana. The last time I checked, using marijuana is a federal crime under a variety of drug laws. Nevertheless, those federal statutes have nothing to do with the patent laws. Please, do not blame patent law for this one.