SmithKline Beecham v. Apotex (on petition for writ of certiorari)
The Supreme Court is contemplating hearing this case that questions the standards for inherent anticipation 
of a patent. The drug in question is Paxil, and the case history has more twists than a Hazzard County road. But, the question now before the Court is simple:
Whether a compound that is inevitably produced by the prior art is inherently anticipated by the prior art, and thus not novel under 35 U.S.C. 102.
Another way to focus on the particular question is whether any human recognition of the inherency is required.
Here, Paxil (a PHC hemihydrate) is inevitably produced by a PHC anhydrate, and PHC anhydrate (A.K.A. Ferrosan) was well known before the Paxil patent.
Judge Posner (sitting by designation) determined that Paxil was not inherently anticipated because its inevitable production from Ferrosan was not a “scientific []certainty.” On appeal, the CAFC reversed that particular decision — finding that the patent was inherently anticipated because PHC hemihydrate “inevitably results from production of” PHC anhydrate.
Now, the Solicitor General has provided the government’s views on the case, and supported the CAFC decision:
The court of appeals correctly held that the patent claim is invalid because PHC hemihydrate was inherently anticipated by the prior art that inevitably produced it. . . . a characteristic of a pre-existing product is not patentable even if no one had previously recognized that characteristic.
Because the Gov’t fully supports the CAFC decision, it also recommended against hearing the case.
- Discussion of the original panel opinion: Part I, Part II, Part III, and part IV.
- Discussion of the en banc order and new appellate opinion: Part V, Part VI.
- My premature conclusion: Part VII.
- Solicitor General’s Brief
- Aaron Barkoff’s Orange Book Blog
Response to D Bucklin: It is not clear what scenario you describe. It appears that you question whether the mutant gene sequence can be patented after it is discovered. I would think that as long as the claim is to the “isolated” mutant sequence, which never had been isolated before, it should be patentable. The SmithKline and Schering decisions could have been different if the “purified” compound had been claimed instead of the compound per se (which is how the compounds were claimed). Same problem in the Ariad case (which must be reversed): the claim was simply directed to a process as it occurs in nature, not limited in some way to an aspect of human intervention. You might want to read In re Crish (Fed Cir 2004)
Response to SF. In my opinion you have hit the nail on the head. The issue of usefulness never came up in the decision in SmithKline. What use is a trace amount of an impurity? Contrast this case with In re Seaborg (CCPA). In that case the claim was directed to “element 95”. The defendants showed that miniscule amounts of element 95 would have been inherently produced during operation of a prior art uranium reactor(“Fermi process”). The CCPA held that this did not rise to the level of anticipation essentially because trace amounts would have been useless and would have contributed nothing to the public benefit.
In Schering v Geneva, claims were directed to DCL (an active metabolite of parent compound loratidine). The Fed Cir invalidated the claim in view of a patent merely disclosing loratidine and suggesting that it be administered to humans. Judge Rader discussed In re Seaborg in the Schering opinion. He distinguished Seaborg by stating that the amounts of element 95 in the Fermi process would have been useless, whereas the amounts of DCL produced when loratidine was administerd are useful.
It is my perception that In both Schering and SmithKline the Court re-wrote Supreme Court and CCPA positions on anticipation when the prior art would literally inherently anticipate but fails to be of any public benefit. The other courts held such “anticipation” to be accidental and of no consequence to patentability of the later useful invention. That precedent could easily have been applied in SmithKline. The facts are pretty square with those of In re Seaborg.
Maybe the Court’s dilemma was that Apotex had already been held to infringe the claim to the hemihydrate because they had trace amounts of it in their anhydrate preps. And to hold infringement without anticipation would disturb the old axiom (what infringes if later would anticipate if earlier). I suppose that an alternative could have been to create a doctrine of “accidental infringement” where the literally infringing activity does not rise to the level of an infringement. This would be similar to a reverse-doctrine of equivalents position.
I think if US Supreme Court affirms the decision of CAFC, then the law regarding inherent anticipation is going to be completely changed. As per today, for Inherent Anticipation, the product must be known and intentional. As in this case, neither the production of PAXIL was known before nor was it intentional. Then, how can it be inherently anticipated is a big question mark!
Minor correction: “Hazzard County” was misspelled in the post – it has 2 z’s.
Thanks.
Of the inevitability issue in the paroxetine case, I had written in Intellectual Property Today in August 2005 (page 7; Imagine: No More Indecision In Intellectual Property Cases):
In the paroxetine case (SmithKline Beecham v.
Apotex, 403 F.3d 1328 (CAFC 2005)), there have been a number of false starts.n11
One reader questioned whether the April, 2005 decision really altered the law of
inherent anticipation, on the basis that it was “inevitable” that paroxetine anhydrate will form the hemihydrate. Although the various judges in the case got bound up in the vanishing polymorph (Vonnegut ice-9) issue presented, one notes the simple chemical reality that an anhydrate contains no water and a hemihydrate contains water (i.e., this is more than a polymorph issue).
It is not inevitable that an anhydrate will convert to a hemihydrate. If the
anhydrate is kept away from water, it will never form the hemihydrate. One of
ordinary skill knows how to keep things away from water (e.g., maintaining
anhydrous ferric trichloride purple, protecting alkali metals, using
titanium tetrachloride). While it may be “natural” for an anhydrate to form a
hemihydrate in the presence of water (driven by the thermodynamics of hydration), it is not inevitable,n12 and, in the absence of water, the conversion won’t happen at all. Presumably, the word “natural” was a
proxy for “water-containing environment.” The idea of “inevitability” may now require
elucidation of conditions.
This will be interesting in view of Schering v Geneva.
CAFC said inherency by anticipation does not require a person of ordinary skill in the art recognize the inherent characterisitics of an invention in the prior art.
Will they completely reverse Schering?
I have some questions as to whether the known/proven utility of the prior art should matter in inherent anticipation and anticipation generally.
Would it matter if a claimed substance was inherently created via a prior art process that had no known utility? For example, suppose a publication described a process to make compound A, but provided no use for compound A. Consequently, the process has no utility. Also, suppose that the process necessarily creates patented compound B in trace amounts, but the publication never mentions compound B. Is it sufficient that the publication discloses a useless process that could make compound B?
What if a publication identified compound B, disclosed a process for making compound B, but provided no use for compound B? The publication would enable one to “make” the compound, but not “make and use” the compound. Should anticipation or inherent anticipation require that some use be shown?
Finally, if “use” doesn’t matter, could someone create a computer program to publish random sequences of compounds and thus foreclose patenting those compounds?
Nucleic acid sequence claims are often directed to a gene or a mutant version.
The disorder caused by the mutation is often known long before the sequence is known. For example, cystic fibrosis was described as a disorder long before the mutations that cause it were known.
In other words, the compound inevitably came before the disorder. But, does the fact that the mutant sequence came before the phenotype save the sequence from inherent anticipation?
The phenotype inevitably follows from the gene sequence. Or, cast in the words of the SG, is a mutant a pre-existing product that no one had previously recognized. If the SG’s position prevails, are discoveries patentable?