Elan Pharm. v. Mayo Foundation for Medical Educ., 175 F. Supp. 2d 1209 (N.D. Cal. 2000)

*1210 Andrew A. Kumamoto, William E. Thomson, Jr., McCutchen Doyle Brown & Enerson, LLP, San Francisco, CA, Lynn H. Pasahow, S. Christian Platt, McCutchen Doyle Brown & Enerson LLP, Palo Alto, CA, for plaintiffs.

Robert E. Hillman, Fish & Richardson P.C., Boston, MA, Karen I. Boyd, Fish & Richardson P.C., Menlo Park, CA, John A. Burtis, Chad A. Hanson, Fish & Richardson, Minneapolis, MN, for defendant.

ALSUP, District Judge.

The issue on defendant's motion for summary judgment is whether a prior-art patent that discloses a "recipe" for producing transgenic mice, some of which will certainly end up with a particular characteristic and some of which won't, anticipates claims based on the same recipe but covering only those mice that end up with the particular characteristic.

A suspected cause of Alzheimer's disease is the formation of plaque in the brain comprised residue from the break-down of an APP polypeptide. This residue is called AΣυ, and the parties agree that it is always produced together with ATF-betaAPP, and vice versa. ATF-betaAPP is not suspected to form plaque. Searching for a cure to Alzheimer's disease, scientists screen various drugs on mice aimed at prohibiting or otherwise affecting the production of mice that produce AΣυ (and therefore also ATF-betaAPP) are useful in this research. The mice that are used in these tests are the patented technology at issue in this case.

*1211 The patents-in-suit are United States Patent No. 5,612,486 and United States Patent No. 5,850,003, assigned to plaintiffs. Plaintiffs have asserted claims 1-3 of the '486 patent and claims 1-3 of the '003 patent. These claims cover rodents harboring a human gene with the Swedish mutation that produce ATF-betaAPP in a sufficient amount to be detectable in a brain homogenate. Independent claim 1 of the '486 patent provides as follows:

A transgenic rodent comprising a diploid genome comprising a transgene encoding a heterologous APP polypeptide having the Swedish mutation wherein the amino acid residues at positions corresponding to positions 595 and 596 in human APP695 are asparagine and leucine, respectively, 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 homogenate of said transgenic rodent.

Independent claim 1 of the '003 patent, parallel to Claim 1 of the '486 patent, provides as follows:

A transgenic rodent comprising a diploid genome comprising a transgene comprising in operable linkage a promoter, a DNA segment encoding a heterologous APP polypeptide and a polyandenylation site, wherein the APP polypeptide has the Swedish mutation whereby the amino acid residues at positions corresponding to positions 595 and 596 in human APP695 are asparagine and leucine, respectively, 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 homogenate of said transgenic rodent.

The remaining claims of each patent depend from these claims.

Not all transgenic rodents harboring a human gene with the Swedish mutation produce ATF-betaAPP in a sufficient amount to be detectable in a brain homogenate. The method disclosed in the patents-in-suit for making mice that produce a detectable amount of ATF-betaAPP is imperfect that is, not all mice made according to the patents' disclosures will produce an amount of ATF-betaAPP detectable in a brain homogenate. Some of the mice made according to the patents' disclosures will not harbor the gene with the Swedish mutation. Of those mice that do harbor the gene, some will not produce the APP polypeptide. Of those that do produce the APP polypeptide, some will not process the APP polypeptide such that ATF-beta APP will be detectable in a brain homogenate.

The earliest date of conception for the claims in the patents-in-suit was June 30, 1993. Defendant has pointed to two references that pre-dated the conception of the claims at issue. The first was an article published in 1992 by Michael Mullan documenting his discovery of an APP mutation in a Swedish family now called the Swedish mutation.

The second was a patent application filed by Mullan in 1992 that disclosed transgenic mice harboring a human APP gene with the Swedish mutation. The Mullan patent application disclosed the same recipe for making transgenic mice as was later disclosed in the patents-in-suit. Mullan disclosed making transgenic mice with the Swedish mutation; he described making constructs to use in making transgenic animals; he described the Swedish mutation at positions 670 and 671 in APP 770 (plaintiffs do not dispute that these are equivalent to positions 595 and 596 in APP695); he described making transgenic mice with the Swedish mutation that express the gene into a mutated protein; and *1212 he described the processing of APP into AΣυ by enzymes. Mullan did not actually make the mice described and claimed in his patent. Mullan did not disclose looking for ATF-betaAPP as a proxy for AΣυ.

In addition to the patents-in-suit, plaintiffs own a patent by the same inventors covering the brain-homogenate method for testing for ATF-betaAPP disclosed in the specifications of the patents-in-suit. That patent has not been asserted against Mayo because Mayo tests its mice directly for AΣυ, as is now the industry standard.

A "party moving for summary judgment bears the burden of identifying the evidence that demonstrates the absence of a disputed material fact and establishes that the moving party is entitled to judgment as a matter of law." See Seal-Flex, Inc. v. Athletic Track And Court Constr., 98 F.3d 1318, 1321 (Fed.Cir. 1996) (citation omitted). "A patent is presumed to be valid, see 35 U.S.C. § 282, and this presumption only can be overcome by clear and convincing evidence to the contrary." Helifix Ltd. v. Blok-Lok, Ltd., 208 F.3d 1339, 1346 (Fed.Cir.2000).

An invention is invalid as anticipated if it "was described in a patent granted on an application for patent by another filed in the United States before the invention thereof by the applicant for patent ...." 35 U.S.C. § 102(e). The prior-art patent must disclose each and every limitation of the claimed invention, must be enabling, and must describe the claimed invention sufficiently to have placed it in possession of a person of ordinary skill in the art. Ibid.

Though each limitation of the asserted claim must be disclosed in the prior-art patent, the prior-art patent need not duplicate the claim word for word. It is sufficient that the claimed limitation is "inherent" in the relevant reference. In re Donohue, 766 F.2d 531, 533 (Fed.Cir.1985). The invention disclosed in the prior-art patent need not have been built or practiced, so long as it is described sufficiently to enable others to build or practice it. Atlas Powder Co. v. E.I. du Pont De Nemours & Co., 750 F.2d 1569, 1576 (Fed.Cir. 1984).

Plaintiffs concede that the Mullan patent, filed in 1992, is prior art to the patents-in-suit. Plaintiffs contend, however, that the Mullan patent does not anticipate the asserted claims because the Mullan patent does not disclose, explicitly or inherently, the transgenic mice processing the APP polypeptide to ATF-betaAPP in a sufficient amount to be detectable in a brain homogenate.^[1] The Court disagrees. 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 Mullan (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 limitation of the final "wherein" clauses of the asserted claims.

Plaintiffs' arguments to the contrary do not persuade the Court. First, plaintiffs argue that all mice created from the Mullan disclosure would have to process the APP polypeptide to ATF-betaAPP in order for Mullan to anticipate the last "wherein" clauses of the asserted claims. Plaintiffs cite Glaxo, Inc. v. Novopharm Ltd., 52 F.3d 1043 (Fed.Cir.1995), to support their position. In Glaxo, Inc., Glaxo *1213 chemists discovered ranitidine in 1976 and began making preparations for manufacturing and selling the compound as an antiulcer drug. Glaxo, Inc. v. Novopharm Ltd., 830 F. Supp. 871, 873 (E.D.N.C.1993). In 1977, Glaxo filed an application for a United States Patent covering all rantidine compounds of a specified general formula, including ranitidine hydrochlorides, a salt of ranitidine that Glaxo had discovered about a month earlier. That application resulted in the '658 patent. The '658 patent disclosed the method, called Example 36, for making rantidine hydrochloride. Ibid.

From 1977 to 1980, Glaxo tested procedures for the production of rantidine hydrochloride. The procedure set out in Example 32 was not used in any of this work. Glaxo used a similar process it called Process 3A, and then later a more efficient process it called Process 3B. In 1980, a batch of rantidine hydrochloride prepared using Process 3B produced crystals that were different from the previously produced by Glaxo. Glaxo determined that there were two crystal forms of rantidine hydrochloride. It called the 1980 variety Form 2 and the earlier variety Form 1. Glaxo eventually decided to proceed with the commercial development of Form 2 rather than Form 1, and modified Process 3B so that it regularly produced Form 2. Ibid. It then applied for and obtained the '431 patent covering Form 2. Id. at 874. It asserted the '431 patent against Novopharm.

As a defense to infringement, Novopharm argued that the '431 patent was inherently anticipated by Glaxo's earlier '658 patent, which disclosed the Example 32 method for making rantidine hydrochloride. Ibid. At trial, Novopharm contended that, practiced as written, the product of Example 32 was always Form 2. Four chemists testified for Novopharm that they had repeatedly followed the Example 32 procedure and always achieved Form 2. Id. at 875. Three chemists for Glaxo, on the other hand, testified that they had repeatedly followed Example 32 and always achieved Form 1. Ibid.

The district court found that the proper practice of Example 32 resulted at times in the exclusive production of Form 1 and at times in the exclusive production of Form 2. From this the court concluded that Novopharm had failed to carry its burden of proving inherence by clear and convincing evidence. According to the court, Form 2 could only have been inherent in Example 32 if Example 32 had invariably produced Form 2 as opposed to Form 1. Ibid. The Federal Circuit affirmed. Glaxo, Inc. v. Novopharm, Ltd., 52 F.3d at 1047-48.

The present case differs significantly in two respects from the Glaxo, Inc. case. First, here, the patents-in-suit do not disclose a new way of making the patented mice, finely tuned to produce only mice that process the APP polypeptide such that ATF-beta APP will be detectable in a brain homogenate. To the contrary, the patents-in-suit disclose the same procedure the only relevant difference is that they claim only a portion of the results. In the Glaxo, Inc., on the other hand, the patent disclosed a new method designed to produce Form 2. Moreover, the disclosure in the prior-art patent indicated that only Form 1 had been achieved, meaning no rantidine hydrochloride in Form 2 had been achieved.

Second, in the present case, a predictable percentage of mice produced according to Mullan (and the patents-in-suit) will inevitably process the APP polypeptide to ATF-beta APP in amounts detectable in a brain homogenate. In Glaxo, Inc., on the other hand, there was nothing inevitable about the production of Form 2 from the Example 32 method, which apparently was a very general instruction. Sometimes *1214 Example 32 produced only Form 1. Sometimes Example 32 produced only Form 2. The results seemed to vary according to how the process was carried out, not according to statistical probabilities. In other words, the practice of Example 32 probably varied in small ways that would unpredictably lead to either Form 1 or Form 2 rantidine hydrochloride.

These differences distinguish the present case from Glaxo, Inc. Here, Mullan's disclosure would produce some mice that process the APP polypeptide to ATF beta-APP in detectable amounts, just as the disclosures in the patents-in-suit produce some mice that process the APP polypeptide to ATF beta-APP in detectable amounts. Were the facts of Glaxo, Inc. parallel to the instant facts, (1) Example 32 would produce a predictable percentage of Form 2 rantidine hydrochloride regardless of who was practicing the procedure, and (2) the '431 patent (the patent there at issue) would not have disclosed a new process for predictably achieving only Form 2, but instead would have disclosed the same Example 32 while claiming only the Form 2 product.

In addition to the Glaxo, Inc. opinions, plaintiffs also cite Finnigan Corp. v. United States Int'l Trade Comm'n, 180 F.3d 1354, 1365 (Fed.Cir.1999). From Finnigan Corp. Plaintiffs quote the following language: "Inherency ... may not be established by probabilities or possibilities. The mere fact that a certain thing may result from a given set of circumstances is not sufficient." But Plaintiffs omit the remainder of the passage. In whole, the passage reads:

Inherency, however, may not be established by probabilities or possibilities. The mere fact that a certain thing may result form a given set of circumstances is not sufficient. If, however, the disclosure is sufficient to show that the natural result flowing from the operation as taught would result in the performance of the questioned function, it seems to be well settled that the disclosure should be regarded as sufficient.

Finnigan Corp., 180 F.3d at 1365 (quoting In re Oelrich, 666 F.2d 578, 581 (Cust. & Pat.App. 1981)). As an initial matter, the Mullan patent meets the above definition of inherency. Mice that produce ATF-betaAPP in detectable amounts are a natural result flowing from the practice taught by Mullan, just as they are a natural result flowing from the (essentially identical) process taught by the patents-in-suit.

In addition, the present case is distinguishable from Finnigan on its facts. In Finnigan, the issue was whether a prior-art article inherently disclosed nonresonance ejection in a quadruple ion trap. At trial, the author of the article was equivocal as to whether the diagrams in his article showed nonresonance ejection or selective ejection. The Federal Circuit held that "[t]he mere possibility that Figure 2 might be understood by one of skill in the art to disclose nonresonnance ejection is insufficient to show that it is inherently disclosed therein." Id. at 1366. The present case does not fit the Finnigan model. Here, there is no ambiguity in the Mullan patent such that the Court must determine whether it would have taught one skilled in the art to make the covered mouse. Rather, here, the parties agree that if one skilled in the art were to follow Mullan, he or she would produce some mice that would process an APP polypeptide having the Swedish to ATF-betaAPP in a sufficient amount to be detectable in a brain homogenate.

Plaintiffs also point to MEHL/Biophile Int'l Corp. v. Milgraum, 192 F.3d 1362, 1365 (Fed.Cir.1999). There, the Federal Circuit found that a prior-art manual that taught how to use a laser to remove tattoos did not inherently disclose a method *1215 of removing hair using a laser because an operator of the tatoo-removing laser could use it without necessarily aligning the laser "substantially vertical over a hair follicle opening." Again, the present case is distinguishable because using the Mullan disclosure (or the disclosures in the patents-in-suit) to make numerous mice, as intended, would inevitably lead to some mice with the characteristics claimed in the patents-in-suit. The mice would not be an accident, but rather a statistical inevitability.

Plaintiffs argue that one could practice Mullan and still not infringe the claims of the patents-in-suit. According to plaintiffs, the claims at issue implicitly require that a successful mouse that is, a mouse with detectable amounts of ATF-betaAPP be identified as such. An untested mouse, according to plaintiffs' logic, could not infringe the patents-in-suit, even if that mouse processed an APP polypeptide having the Swedish to ATF-betaAPP in a sufficient amount to be detectable in a brain homogenate. The Court disagrees with this position. The asserted claims do not require testing of the covered mouse. They require only that the mouse have claimed characteristics.

For the foregoing reasons, the Court holds that the Mullan patent anticipates the asserted claims of the patents-in-suit. Defendant's motion for summary judgment is GRANTED. The clerk shall close the file.

IT IS SO ORDERED.

[1] Plaintiffs do not argue that the asserted dependent claims avoid Mullan by reason of limitations not found in the independent claims. The following analysis therefore applies to all asserted claims.