dissenting.
The majority decision in this case rests upon the ground that an inventor’s own disclosure may not be used under 35 U.S.C. § 102 as proof of anticipation by inherent disclosure in a prior art reference. This decision contradicts our own case law, which holds that knowledge of an inherent characteristic in the prior art is irrelevant. As we recently recognized in In re Cruciferous Sprout Litigation, 301 F.3d 1343, 1350 (Fed.Cir.2002), on the issue of inherency “[i]t matters not that those of ordinary skill heretofore may not have recognized these inherent characteristics.” Here, as in Cruciferous, while Elan “may have recognized something quite interesting about those [mice], it simply has not invented anything new.” Id. at 1351. This decision, if followed, will have serious and unfortunate consequences in the future by permitting the securing of patent rights to existing inventions so long as the patent applicant identifies an inherent characteristic of that product that was not identified in the prior art. That has never been our law. I respectfully dissent.
I
The patents asserted herein are U.S. Patent Nos. 5,612,486 (“the '486 patent”) and 5,850,003 (“the '003 patent”) (collectively “the Elan patents”). The sole independent claim of the '486 patent recites in relevant part:
A transgenic rodent ... comprising a transgene encoding a heterologous APP *1232polypeptide having the Swedish mutation ... wherein the transgene is expressed to produce a human APP polypeptide having the Swedish mutation, and wherein said polypeptide is processed to ATF-betaAPP in a sufficient amount to be detectable in a brain hom-ogenate of said transgenic mouse.
Claim 1 of the '486 patent (emphases added). The sole independent claim of the '003 patent recites in relevant part:
A transgenic rodent ... comprising a transgene comprising ... a DNA segment encoding a heterologous APP polypeptide ..., wherein the transgene is expressed to produce a human APP polypeptide having the Swedish mutation, and wherein said polypeptide is processed to ATF-betaAPP in a sufficient amount to be detectable in a brain hom-ogenate of said transgenic rodent.
Claim 1 of the '003 patent (emphases added). Because these claims are directed to transgenic rodents, the methods by which they are produced are not elements of the claims. Nor is there any claim to a method for detecting ATF-betaAPP.
Despite the clear language of the claims mandating their interpretation as products (transgenic rodents), much effort both at the district court and here on appeal has been expended on arguments incorrectly interpreting the claims in terms of methods. Elan argues, for example, that U.S. Patent No. 5,455,169 to Mullan (hereinafter “Mullan”), cited by the Mayo Foundation for Medical Education and Research (hereinafter “Mayo”) as anticipating the claims of the Elan patents, fails to teach “how to detect ATF-betaAPP, much less how to detect the fragment in a brain homogenate.” (Appellants’ Br. at 23.) The claims of the Elan patents, however, require only detectable ATF-betaAPP and not a method of detection.
According to Elan, the '486 and '003 “patents required that its transgenic mice do all these things: [1] carry the APPSw transgene, [2] express the APPSW protein and [3] process the APPSW to ATF-be-taAPP such that the levels of ATF-be-taAPP are detectable.” (Appellants’ Br. at 21.) As admitted by Elan in its brief on appeal, “Elan does not dispute that the specification of the Mullan patent disclosed a transgenic mouse harboring a human APP gene with the Swedish mutation.” Id. at 17. In other words, the first element was disclosed. On appeal Elan also does not contend that the second element was not disclosed.1 Elan contests solely the third aspect of the, claims. Elan bases the novelty of its claimed rodents on the “critical element — processing APP to ATF-betaAPP in an amount sufficient to be detectable in a brain homogenate.” Id. Mayo concedes that Mullan fails to expressly disclose this element of the claimed invention, but counters that this characteristic was inherent in the disclosure of Mul-lan. The only issue, therefore, is whether *1233the rodent of Mullan will inherently produce ATF-betaAPP in a sufficient amount to be detectable in its brain homogenate.
II
On summary judgment the district court ruled that Mullan inherently anticipates the claims of the Elan patents, finding:
The mice claimed in the patents-in-suit are merely a subset of the mice described in Mullan. Some of the mice made using the process disclosed in Mul-lan (which is essentially the same process disclosed in the patents-in-suit) would inevitably have detectable levels of ATF-betaAPP. Were Plaintiffs to contend otherwise, their own patents would not be enabled. Mullan therefore inherently includes the [detectable ATF-betaAPP] limitation of the final “wherein” clauses of the asserted claims.
Elan Pharms., Inc., 175 F.Supp.2d at 1212.
The majority disagrees, apparently because no extrinsic evidence of inherency existed in the prior art. The majority states:
there was no evidence that the formation and detection of ATF-betaAPP in the transgenic mouse brain with the Swedish mutation was known to persons of ordinary skill in the field of the invention. Inherency cannot be based on the knowledge of the inventor; facts asserted to be inherent in the prior art must be shown by evidence from the prior art. Cf. In re Dembiczak, 175 F.3d 994, 999, 50 USPQ2d 1614, 1617 (Fed.Cir.1999) (criticizing the “hindsight syndrome wherein that which only the inventor taught is used against its teacher”).
Ante at 1228.
But this is not the correct analysis. This is not an obviousness case. The injunction in Dembiczak against using an inventor’s own disclosure against him was in the context of a section 103 obviousness determination requiring “the oft-difficult but critical step of casting the mind back to the time of invention, to consider the thinking of one of ordinary skill in the art, guided only by the prior art references and the then-accepted wisdom in the field.” 175 F.3d at 999, 50 USPQ2d at 1617 (citing W.L. Gore & Assocs., Inc. v. Garlock, Inc., 721 F.2d 1540, 1553, 220 USPQ 303, 313 (Fed.Cir.1983), cert. denied, 469 U.S. 851, 105 S.Ct. 172, 83 L.Ed.2d 107 (1984)). The perceived problem with combining references using hindsight to render a claimed invention obvious is that it “simply takes the inventor’s disclosure as a blueprint for piecing together the prior art to defeat patentability.” Id. This fear of hindsight recreation in the context of obviousness determinations, however, is not applicable in the context of inherency.
There is simply no basis in our law to support the proposition that the source of proof for inherency must be found in the prior art and cannot be found in a paten-tee’s own disclosure or other source. In Continental Can Co. USA v. Monsanto Co., 948 F.2d 1264, 20 USPQ2d 1746 (Fed.Cir.1991), the court noted that “[t]o serve as an anticipation when the reference is silent about the asserted inherent characteristic, such gap in the reference may be filled with recourse to extrinsic evidence.” Id. at 1268, 948 F.2d 1264, 20 USPQ2d at 1749. Thus evidence extrinsic to the cited prior art reference may be used, i.e., the party raising the issue of inherency may fill in the gap in the disclosure using any source. The majority’s contrary conclusion is incorrect as a matter of law, and directly contradicts our law, which has repeatedly recognized that the discovery of an inherent characteristic of an old product cannot be patented. Cruciferous, at 1350; In re Spada, 911 F.2d 705, 709, 15 *1234USPQ2d 1655, 1658 (Fed.Cir.1990) (“When the claimed [inventions] are not novel they are not rendered patentable by recitation of properties, whether or not these properties are shown or suggested in the prior art.” (emphasis added)); Titanium Metals Corp. of Am. v. Banner, 778 F.2d 775, 782, 227 USPQ 773, 779 (Fed.Cir.1985) (“[I]t is immaterial, on the issue of their novelty, what inherent properties the [disclosed products] have or whether these applicants discovered certain inherent properties”).
Ill
Because the disclosures of the Elan patents may be used as proof that the Mullan transgenic rodent inherently possessed the claimed characteristic, the remaining question is whether the Elan patents, in fact, provide that proof. They clearly do. The specification of the '003 patent teaches:
Newly identified secreted fragments comprise amino-terminal portion of JAPP (Ap) which remains after the cleavage and will be referred to hereinafter as the amino-terminal fragment form of pAPP (ATF-pAPP) [ATF-be-taAPP], ATF-pAPP is believed to be the product of an alternative secretory processing pathway for Ap, which pathway is present even in normal (non-diseased) cells. It is further believed, however, that the alternate secretory pathway may be responsible for an essential event in the production of Ap in diseased cells in patients, and that abnormal production of AFT-pAPP may be involved in diseases related to Ap plaque....
Particularly preferred animal models for p-secretase cleavage of Ap are transgenic animals which express the Swedish mutation of the Ap gene.... It has been found that such transgenic animals, particularly transgenic mice, produce high quantities of the AFT [sic ATF]-$APP which may [be] detected according to the methods of the present invention. In particular, it has been found that Swedish mutation o/Ap produces quantities of the ATF-&APP which will usually be at least two-fold higher than unid type human pAPP expressed in animals.
'003 patent, col. 12, ll. 21-42 (emphases added). The “discoveries” discussed in the preceding passage are two-fold: first, that the P-secretase cleavage (metabolism) of the Swedish mutation form of APP to produce the p-amyloid peptide (pA) results in a secondary “newly identified” fragment, ATF-pAPP; and second, that the newly discovered fragment is found in “high quantities” in transgenic mice having the Swedish mutation form of APP.
As Elan concedes on appeal, “the specification of the Mullan patent disclosed a transgenic mouse harboring a human APP gene with the Swedish mutation.” (Appellants’ Br. at 17.) More than simply “harboring” the gene as suggested by Elan, however, Mullan discloses a transgenic mouse that will express the gene to produce the Swedish APP and then metabolize the APP to produce the p-amy-loid peptide for the study of the underlying biochemistry of that metabolism. Mullan, col. 11, ll. 5-36 (“[S]uch model systems provide a tool for defining the underlying biochemistry of APP and p-amyloid metabolism.... The human variant APP codon ... can be transferred to a host non-human, animal, such as a mouse. As a result of the transfer, the resultant transgenic non-human animal will express one or more variant APP co-don 670/1 polypeptides.”). As disclosed in the '003 specification, Swedish APP to p~ amyloid metabolism directly produces the “newly identified” ATF-pAPP metabolite. '003 patent, col. 12, 11. 21-22. Further, transgenic mice that carry out the Swed*1235ish APP to p-amyloid metabolism produce “high quantities” of the ATF APP metabolite. Id. at col. 12, 11. 35-42. Because the claims are not limited to a particular “method of detection,” but rather broadly recite the requirement that the fragments be “detectable,” a mouse that metabolizes APP to produce the (3-amyloid peptide in sufficient amounts to permit the study of the underlying biochemistry of that metabolism would necessarily produce detectable amounts of the ATF-(3APP metabolite.
Elan argues that “[t]he transgenic mice claimed by Elan’s patents are only a tiny, and at the time of Mullan unexpected, subset of the larger population of transgenic mice that might be produced by following the Mullan ‘recipe.’” (Appellants’ Reply Br. at 6.) In fact, the claimed mice are not a tiny subset of the mice disclosed in Mullan. To be sure, Mullan discloses a method for producing transgenic mice not all of which will successfully express the Swedish form APP. However, the Swedish form APP characteristic is disclosed in Mullan, and in each and every case of a mouse that processes Swedish form APP to produce the (3-amyloid chain as disclosed in the Mullan patent, that mouse will also produce ATF-(3APP as claimed in the Elan patents. '003 patent, col. 12,11. 21-42. Thus, the rule that “[in-herency ... may not be established by probabilities or possibilities,” In re Oelrich, 666 F.2d 578, 581, 212 USPQ 323, 326 (CCPA 1981) (citation omitted), is not violated by finding the claims of the Elan patent anticipated by a mouse according to Mullan (expressing and metabolizing the Swedish form APP), which will always possess the ATF-pAPP characteristic.
The district court correctly concluded that the claims of the Elan patents are invalid as inherently anticipated by Mul-lan.
For the foregoing reasons, I respectfully dissent.
. The majority appears to suggest that this element was not disclosed in Mullan, but this issue was not raised on appeal. In any event, the Mullan patent discloses the second element, stating "[a]s a result of the transfer, the resultant transgenic non-human animal will express one or more variant APP codon 670/1 polypeptides.” Mullan, col. 11, 11. 34-36. The majority cites no authority suggesting that any more detailed description was required. To be sure the Mullan reference must have been enabling in this respect, In re Donohue, 766 F.2d 531, 533, 226 USPQ 619, 621 (Fed. Cir.1985), and there may be a question as to whether it was enabling. But Elan has deliberately decided not to mount,an enablement challenge to the Mullan patent, apparently for the reasons explained by the district court • relating to potential for such arguments to invalidate Elan’s own claims for lack of en-ablement. Elan Pharms., Inc. v. Mayo Found, for Med. Educ. & Research, 175 F.Supp.2d Í209, 1212 (N.D.Cal.2000).