Therasense, Inc. v. Becton, Dickinson and Co.

Opinion for the court filed by Circuit Judge DYK. Opinion concurring-in-part and dissenting-in-part filed by Circuit Judge LINN.

DYK, Circuit Judge.

Therasense, Inc. (now Abbott Diabetes Care, Inc.) and Abbott Laboratories appeal from a final judgment of the United States District Court for the Northern District of California. Following a bench trial, the district court determined that claims 1-4 of U.S. Patent No. 5,820,551 (“the '551 patent”) were invalid due to obviousness and that the entire '551 patent was unenforceable due to inequitable conduct. See Therasense, Inc. v. Becton, Dickinson & Co., 565 F.Supp.2d 1088 (N.D.Cal.2008) (“Trial Opinion ”). The district court also granted summary judgment of noninfringement of all asserted claims in U.S. Patent Nos. 6,143,164 (“the '164 patent”) and 6,592,745 (“the '745 patent”). See Therasense, Inc. v. Becton, Dickinson & Co., 560 F.Supp.2d 835 (N.D.Cal.2008) (“Summary Judgment Order”). Finally, it found nearly all of the asserted claims of the '745 patent invalid due to anticipation. Id. We affirm.

BACKGROUND

I

The Centers for Disease Control and Prevention estimates that 23.6 million Americans have diabetes. Ctrs. for Disease Control & Prevention, National Diabetes Fact Sheet, 2007, at 5 (2008). Diabetes mellitus, generally referred to simply as diabetes, is a disorder of the carbohydrate metabolism in which the body either does not produce enough or does not properly utilize insulin, a hormone produced in the pancreas. Insulin regulates blood sugar levels by facilitating the absorption of glucose into cells. In diabetics, this failure of the body to produce or utilize insulin causes glucose to build up in the blood, leading to potentially serious complications, such as comas, blindness, kidney damage, cardiovascular disease, and lower-limb amputations. Diabetics today rely on inexpensive and accurate home blood glucose meters to help them manage their condition.

The patents-in-suit claim technology in the area of disposable blood glucose test strips. These single-use strips work in conjunction with glucose meters and employ electrochemical biosensors to measure the level of glucose in a sample of blood, usually a single drop. When blood is introduced to a test strip, the glucose in the blood reacts with an enzyme present on the strip, transferring electrons to the enzyme. A mediator, also incorporated on the strip, transfers the electrons from the enzyme to the sensor’s “active electrode.” The electrons then flow from the strip into the meter, which calculates the glucose concentration based on the current flow. While common test strips measure glucose levels in samples of whole blood (blood that contains all of its components, including red blood cells), electrochemical sensors are also capable of measuring glucose levels in a variety of other sample types, such as blood plasma (blood without red blood cells), interstitial fluid (fluid between cells in the skin and organs), and buffer solutions (test solutions that simulate blood). They are also capable of measuring samples in vivo—inside the body—such as live blood. We describe below the various patents involved here.

*1293II

Abbott Diabetes Care, Inc., the successor to Therasense, Inc. and a subsidiary of Abbott Laboratories, is the owner of the patents-in-suit. For convenience we will use the single designation “Abbott” to refer to any and all of the foregoing. In March 2004, Becton, Dickinson and Co. (“BD”) sued Abbott in the United States District Court for the District of Massachusetts seeking a declaratory judgment of noninfringement of the '164 and '745 patents. The two products involved were the BD Latitude Diabetes Management System and the BD Logic Blood Glucose Monitor. Both products featured BD’s blood glucose test strip, the BD Test Strip. Two months later, Abbott countersued BD in the Northern District of California alleging infringement of the '164, '745, and '551 patents. The action in the District of Massachusetts was subsequently transferred to the Northern District of California. Shortly thereafter, Abbott also initiated an infringement action against Nova Biomedical Corp. (“Nova”), BD’s supplier, alleging infringement of the '164, '745, and '551 patents asserted in Abbott’s suit against BD.1 In August 2005, Abbott sued Bayer Healthcare LLC (“Bayer”) alleging that its Microfill and Autodisc blood glucose strips infringed the '551 and '745 patents.

All of the cases were consolidated, with the defendants asserting that their products did not infringe Abbott’s patents and that the patents were invalid. On April 3, 2008, the court issued a summary judgment order finding that BD/Nova’s products did not infringe any of the asserted claims of the '164 or '745 patents. It also found claims 1-5, 8, 21-23, 28, 31, and 34 in the '745 patent anticipated. Summary Judgment Order, 560 F.Supp.2d at 880. Issues relating to the '551 patent were considered in a bench trial from May 27-June 3, 2008. After trial the district court determined that claims 1-4 of the '551 patent were invalid due to obviousness and that the entire '551 patent was unenforceable due to inequitable conduct. Trial Opinion, 565 F.Supp.2d at 1127.2 The district court entered Rule 54(b) judgments, and we have jurisdiction over this appeal pursuant to 28 U.S.C. § 1295(a)(1).

DISCUSSION

I Obviousness of the '551 Patent

We first address the district court’s holding that claims 1-4 of the '551 patent would have been obvious in light of U.S. Patent Nos. 4,545,382 (“the '382 patent”) and 4,225,410 (“the '410 patent”). Obviousness is a question of law based on underlying questions of fact. Daiichi Sankyo Co., Ltd. v. Apotex, Inc., 501 F.3d 1254, 1256 (Fed.Cir.2007) (citing Winner Int’l Royalty Corp. v. Wang, 202 F.3d 1340, 1348 (Fed.Cir.2000)). We therefore review the ultimate determination of obviousness by a district court de novo and the underlying factual inquiries for clear error. Id.

A Membraneless Sensor

The claims of the '551 patent describe a test strip with an electrochemical sensor *1294for testing whole blood without any membrane over the electrode.3 Some sensors in the prior art had employed diffusion-limiting membranes to control the flow of glucose to the electrode because the slower mediators of the time could not deal with a rapid influx. Other prior art sensors employed protective membranes to prevent “fouling”—when red blood cells stick to the active electrode and interfere with electron transfer to the electrode, resulting in an inaccurate measurement. In addition, sensors injected into the human body for in vivo measurements used protective membranes as a safety measure to prevent the chemistry from dissolving into the body and because fouling was a particular problem with respect to long-term in vivo implants.

The central question with respect to obviousness is whether the prior art disclosed a glucose sensor without a membrane for use in whole blood. The district court found after trial that the prior art '382 patent disclosed electrochemical sensors in which “a protective membrane was optional in all cases except the case of live blood, in which case the protective membrane was preferred—but not required.” Trial Opinion, 565 F.Supp.2d at 1103. Whether the '382 patent disclosed a membraneless sensor is a question of fact which we review for clear error. See Graham v. John Deere Co., 383 U.S. 1, 17, 86 S.Ct. 684,15 L.Ed.2d 545 (1966).

Initially filed in 1981 by one of Abbott’s predecessors, the '382 patent represented an early achievement in the field of electrochemical sensors for measuring glucose levels. The '382 patent disclosed a faster-acting ferrocene mediator chemistry that allowed for better response times and quicker test results. The first two named inventors on the '382 patent are also the first two named on the '551 patent.

At the outset, it is important to understand the scope and context of the *1295'382 patent. The claims of a prior art patent are part of its disclosure. In re Benno, 768 F.2d 1340, 1346 (Fed.Cir.1985) (“[I]t is true ... that ‘a claim is part of the disclosure’.... ”); In re Smolak, 24 C.C.P.A. 1132, 88 F.2d 838, 841 (1937) (“[T]he disclosures in [prior art] specifications ... include the claims, the written specification, and the drawings.” (quotation omitted)); see also Gabrielidis v. Prince Sports Group, Inc., Nos. 99-1469, 99-1490, 2000 WL 1648134, at *7 (Fed.Cir. Nov. 1, 2000). The claims of the '382 patent are plainly directed in part to sensors without a membrane, as is made clear by the dependent claims that specifically include a membrane as an additional feature of the device. For example, claim 1 claims a sensor electrode utilizing an enzyme and a ferrocene mediator for generating an electrical current representative of the level of a substance in a sample liquid, see '382 patent col. 10 11.52-63, while dependent claim 12 adds “an outermost protective membrane permeable to water and glucose molecules, said membrane covering said enzyme located upon said ferrocene layer,” id. col.ll 11.29-32. One embodiment described in the patent involves a membrane-less sensor for testing interstitial fluid, see id. col.3 1.53-eol.41.2, and Example 8 of the patent specifically describes tests on strips using a buffer solution both with and without membranes, see id. col.9 11.22-30.

Abbott appears not to contest that the '382 patent broadly claims membraneless strips and that the specification discloses such membraneless strips. It contends, however, that in the case of blood—as opposed to other fluids—a membrane is required. In this respect it is significant that none of the claims in the '382 patent explicitly distinguishes between blood and other fluids. Also pertinent is the difference between in vivo sensors and in vitro sensors. In vivo sensors test live blood or other fluids inside the body, while in vitro sensors test fluids such as blood after they have been extracted from the body. Claims 1-17 of the '382 patent cover both in vivo and in vitro sensors; additional claims 18-19 are limited to in vivo sensors, though again those claims do not distinguish between sensors for blood and other fluids.

Thus, as the district court found, the difficulty with Abbott’s position that membranes are required for blood testing is that the claims covering membraneless sensors do not exclude blood and that other fluids are described as being tested by sensors without membranes. Most significantly, the specification addresses blood directly, stating that:

Optionally, but preferably when being used on live blood, a protective membrane surrounds both the enzyme and the mediator layers, permeable to water and glucose molecules.

Id. col.4 11.63-66. On its face, the specification appears to contradict Abbott’s position. It states that when testing with blood—blood in the body (i.e., in vivo testing)—a protective membrane is preferred. The use of the term “preferably” implies that such a membrane is not necessary. See, e.g., Callaway Golf Co. v. Acushnet Co., 576 F.3d 1331, 1346-47 (Fed.Cir.2009) (holding that despite stating that the outer layers of a golf ball were “preferably” made of ionomer resins, the patent also disclosed the use of other materials such as polyurethane); Halliburton Energy Servs., Inc. v. M-I LLC, 514 F.3d 1244, 1251 (Fed.Cir.2008) (“[T]he specification states that ‘preferably’ none of these clays are added; this strongly suggests that absence of clays is simply a preferred embodiment.”); Cordis Corp. v. Medtronic AVE, Inc., 339 F.3d 1352, 1357 (Fed.Cir.2003) (“[U]se of the term ‘preferably’ makes clear that the language describes a preferred embodiment, not the invention *1296as a whole.”). Furthermore, when testing other sample types—which include whole blood outside the body, as well as blood plasma, interstitial fluid, and buffer solutions—the protective membrane is “optional,” even more strongly confirming that it is not required for whole blood. The language of the patent thus confirms that a membrane is not necessary when testing live blood in vivo or whole blood in vitro.

Abbott, however, contends that a person having ordinary skill in the art would not have read the '382 specification that way. Abbott asserts that the conventional wisdom of those skilled in the art was that a membrane was necessary when testing with blood, and that skilled artisans would not have read the patent’s disclosure literally when it said that a membrane was not necessary with blood.

Abbott may well be correct that for in vivo blood testing—-where the sensor is implanted in the body for up to a year—a membrane might sometimes be required for safety and for accurate measurement because of the risk of fouling over the long period of time that the sensor must operate.4 The question, however, is whether a membrane was required for in vitro testing of whole blood where the blood stays on the sensor for, at most, a matter of minutes. Dr. Johnson, Abbott’s expert, testified that one of ordinary skill in the art in 1983 would have thought that a membrane was required with blood with the '382 patent because of the risk of fouling. He did not specifically address the differences between in vivo and in vitro testing of blood. Bayer’s expert witness, on the other hand, specifically addressed both types of testing, recognizing that for in vivo testing of blood a membrane might be necessary to ensure patient safety from the risk of chemicals escaping the sensor and entering the body, and to eliminate the problem of fouling that would occur as the sensor remained in the body for weeks or months. He testified that with in vitro testing a membrane was not necessary because there was no risk of chemicals entering the body and the time period involved in the testing was very short, particularly with respect to the faster chemistry claimed in the '382 patent. The district court plainly found the Abbott testimony not credible and credited the Bayer testimony, stating:

Skilled artisans would have known that deleting the membrane would simply have deleted their mechanical advantages. They would have known, however,- that the electrochemistry would still have worked. They would have known that the degree of fouling would have depended on how long the sensor was exposed to blood. They would have known that the risk of fouling would have been reduced for faster-acting chemistry and reduced even more for sensors used only once, i.e., disposable sensors with no accumulation of residue. They would have known that omitting the filter would have had the further advantage of speeding up the test time even more.

Trial Opinion, 565 F.Supp.2d at 1101. The district court did not clearly err in crediting Bayer’s expert testimony.

Abbott also contends that another portion of the '382 patent’s specification beyond the “[ojptionally, but preferably” sentence demonstrates that a membrane is required when testing whole blood. Abbott in particular points to Example 8 of the patent. Abbott theorizes that the fail*1297ure of Example 8 (listing test results for a buffer solution with and without a membrane and blood with a membrane) to mention whole blood testing without a membrane somehow demonstrates that whole blood testing required a membrane. We reject this argument. A single example testing whole blood with a membrane does not imply that membranes are required. In fact, there was testimony from Dr. Turner that Example 8’s tests with the buffer solution demonstrated that the membrane-less sensor would have worked with blood and provided a suggestion to do so, because the pH level of the buffer was apparently deliberately chosen in this example to match the pH of blood. The district court correctly found that “Example 8 in the '382 patent [is] consistent with the plain meaning of [the ‘[o]ptionally, but preferably’] sentence.” Trial Opinion, 565 F.Supp.2d at 1100. The court thus did not clearly err in finding that the '382 patent disclosed a membraneless sensor for whole blood testing.5

B Enablement and Reasonable Expectation of Success

In order to render a claimed apparatus or method obvious, the cited prior art as a whole must enable one skilled in the art to make and use the apparatus or method. Beckman Instruments, Inc. v. LKB Produkter AB, 892 F.2d 1547, 1551 (Fed.Cir.1989). An individual prior art reference, on the other hand, “need not be enabled; it qualifies as a prior art, regardless, for whatever is disclosed therein.” Amgen, Inc. v. Hoechst Marion Roussel, Ina, 314 F.3d 1313, 1357 (Fed.Cir.2003); see also Symbol Techs., Inc. v. Opticon, Inc., 935 F.2d 1569, 1578 (Fed.Cir.1991); Beckman Instruments, 892 F.2d at 1551.

Abbott contends that the '382 patent in combination with the '410 patent could not enable the membraneless sensor of the '551 invention because the oxygen sensitivity of the '382 patent renders it ineffective when testing whole blood without a membrane. Oxygen sensitivity is an indicator of the effect that an oxygen-rich sample will have on the ability of a sensor to provide an accurate reading. Example 8 of the '382 patent, for example, notes that tests measuring glucose levels in an oxygen-saturated buffer solution returned a reading 95% of that produced in an oxygen-depleted buffer. '382 patent col.9 11.19-21. The '382 patent thus displayed a 5% oxygen sensitivity. Both sides’ experts agreed that a 5% oxygen sensitivity is tolerable.6 The dispute between Abbott’s expert and the defense experts was whether testing with whole blood would lead to oxygen sensitivity of greater than 5%.

The district court made no explicit finding that the '382 patent would have enabled one having ordinary skill in the art to make and use a membraneless sensor. The district court did find, however, that the '382 patent exhibits “very low oxygen sensitivity” comparable to that disclosed in the '551 patent, and therefore the '382 patent would not have any issues with oxygen sensitivity. See Trial Opinion, 565 F.Supp.2d at 1104. This finding is not clearly erroneous. Although there was *1298conflicting testimony on this issue, the district court could properly credit Bayer’s expert, Dr. Turner, who testified that Example 8 of the '382 patent showed that there would be no greater than 5% oxygen sensitivity when testing with whole blood because the buffer test showed only 5% oxygen sensitivity and whole blood would necessarily have a lower oxygen sensitivity than the buffer solution. In light of the district court’s finding that the '382 patent exhibits “very low oxygen sensitivity” comparable to that of the '551 patent, we reject Abbott’s contention that the oxygen sensitivity of the '382 patent does not enable a membraneless sensor.

Abbott also argues that even if the '382 patent disclosed a membraneless sensor for whole blood, the '551 patent would not have been obvious because the '382 patent’s disclosure did not provide a reasonable expectation of success. Here, Abbott essentially recycles its contentions regarding the conventional wisdom of those skilled in the art concerning the need for a membrane and the oxygen sensitivity issue.7 For the reasons discussed above rejecting those contentions, Abbott’s claims that there was no reasonable expectation of success similarly fail.

C Motivation to Combine Prior Art References

The district court found that one skilled in the art would have “readily thought to combine” the '382 patent and the '410 patent to create the membraneless sensor for whole blood of the '551 patent. Trial Opinion, 565 F.Supp.2d at 1123. While the '382 patent discloses a membraneless sensor, the '410 patent teaches the use of a disposable strip-type electrode cartridge that attaches to readout circuitry. Together, these disclose every element of the '551 patent’s blood glucose test strip. Abbott argues that there was no motivation to combine these references because persons of skill in the art would not have “combined a sensor that they did not believe would work in whole blood with the other components to produce a ‘seemingly inoperative’ test strip for testing whole blood.” Pls.-Appellants’ Br. 37-38.

Abbott’s characterization of the '382 patent’s disclosure of a membraneless sensor as “seemingly inoperative” simply repeats its arguments related to enablement and reasonable expectation of success. In light of our conclusions that the district court did not clearly err in finding that the '382 patent disclosed a membraneless sensor and that those with ordinary skill in the art would have expected such a membraneless sensor to work, we conclude that the district court did not clearly err in finding that those with ordinary skill would have been motivated to combine this sensor in a disposable form (the contribution of the '410 patent). Dr. Turner testified that the field of biosensors in the early 1980s was a “sophisticated field.” J.A. 2523. He also testified that at the time, persons with ordinary skill in the art recognized the advantages to the strip-type format for home testing, which would have motivated a person to put the membrane-less sensor on a strip. Abbott failed to present any expert evidence to rebut that expert testimony. Therefore, the district court did not clearly err in finding that a motivation to combine existed.

*1299D Secondary Considerations: Commercial Success

Abbott contends that even if the district court were correct in finding a prima facie case of obviousness, that finding is overcome by evidence of commercial success. The district court concluded that Abbott failed to demonstrate a sufficient nexus between the claims of the '551 patent and the success of Abbott’s Exactech product. Trial Opinion, 565 F.Supp.2d at 1124. This finding is not clearly erroneous.

In order to overcome a finding of obviousness by demonstrating commercial success, “[a] nexus between commercial success and the claimed features is required.” Brown, & Williamson Tobacco Corp. v. Philip Morris Inc., 229 F.3d 1120, 1130 (Fed.Cir.2000). “[T]he asserted commercial success of the product must be due to the merits of the claimed invention beyond what was readily available in the prior art.” J.T. Eaton & Co. v. Atl. Paste & Glue Co., 106 F.3d 1563, 1571 (Fed.Cir.1997).

Abbott argues on appeal that the district court erred by requiring Abbott to show that the commercial success of the Exactech product was attributable to the '551 patent, and that Abbott was entitled to a presumption that the commercial success was due to the invention claimed in the '551 patent. Abbott claims that “if the marketed product embodies the claimed features, and is coextensive with them, then a nexus is presumed and the burden shifts to the party asserting obviousness to present evidence to rebut the presumed nexus.” Brown & Williamson, 229 F.3d at 1130.

Abbott is incorrect in contending that it was entitled to the presumption of a nexus. This is not a situation where the success of a product can be attributed to a single patent, because Abbott’s Exactech product embodied at least two patents: the '382 patent and the '551 patent. Abbott’s expert, Dr. Johnson, admitted that the Exactech strips met all of the limitations of claim 1 of the '382 patent. Furthermore, for the fifteen years that the product was on the market—during which the '382 patent was valid and in force for the entire period save for the final three months— Abbott marked the product with the '382 patent, both before and after the '551 patent issued. As such, there is no presumption that the product’s success was due only to the '551 patent.

The defendants presented uncontroverted evidence demonstrating that the Exactech product’s success was due to features already present in prior art such as the '382 patent. Abbott’s primary witness on commercial success conceded that the Ex-actech product was successful because it offered “fast disposal for whole blood that could directly be put onto the strip, [provided] the reading, and then basically that strip [was] disposable.” J.A. 2640. In addition, the defendants presented numerous declarations that Abbott had filed with the PTO during the prosecution of the '551 patent describing the success of the Exactech strips, all of which attested to the product’s “ease of use, short user training period, and ability to consistently obtain accurate test results,” J.A. 6969, and “[t]he electrode configuration (allowing a small sample size), the size of the test strips, and the ease with which they can be stored, used and disposed of,” J.A. 6985. Abbott offered nothing to contradict this evidence and show that the commercial success of the Exactech product was due to the lack of a protective membrane over the sensor electrode. Under these circumstances, the district court did not err in rejecting Abbott’s assertion that commercial success supported a finding of nonobviousness.

For the reasons stated above, we find no error with the district court’s conclusion *1300that claims 1-4 of the '551 patent would have been obvious in light of the prior art.

II Inequitable Conduct in the Prosecution of the '551 Patent

Following the bench trial, the district court also held the '551 patent unenforceable for inequitable conduct based on a failure to disclose statements made to the European Patent Office (“EPO”) during a revocation proceeding of the European counterpart to the '382 patent. Trial Opinion, 565 F.Supp.2d at 1127.

“[Unequitable conduct includes affirmative misrepresentation of a material fact, failure to disclose material information, or submission of false material information, coupled with an intent to deceive.” Innogenetics, N.V. v. Abbott Labs., 512 F.3d 1363, 1378 (Fed.Cir.2008) (quoting Pharmacia Corp. v. Par Pharm., Inc., 417 F.3d 1369, 1373 (Fed.Cir.2005)) (quotation marks omitted). “The party asserting inequitable conduct must prove a threshold level of materiality and intent by clear and convincing evidence. The court must then determine whether the questioned conduct amounts to inequitable conduct by balancing the levels of materiality and intent, “with a greater showing of one factor allowing a lesser showing of the other.’ ” Digital Control, Inc. v. Charles Mach. Works, 437 F.3d 1309, 1313 (Fed.Cir.2006) (quoting Union Pac. Res. Co. v. Chesapeake Energy Corp., 236 F.3d 684, 693 (Fed.Cir.2001)) (citations omitted).

The ultimate finding of inequitable conduct is reviewed under an abuse of discretion standard. Kingsdown Med. Consultants, Ltd. v. Hollister Inc., 863 F.2d 867, 876 (Fed.Cir.1988) (en banc). We review the underlying factual determinations of materiality and intent for clear error, however. Star Scientific, Inc. v. R.J. Reynolds Tobacco Co., 537 F.3d 1357, 1365 (Fed.Cir.2008).

The penalty for inequitable conduct is severe, as an entire patent is rendered unenforceable. Therefore it is important that courts maintain a high standard. “Just as it is inequitable to permit a patentee who obtained his patent through deliberate misrepresentations or omissions of material information to enforce the patent against others, it is also inequitable to strike down an entire patent where the patentee only committed minor missteps or acted with minimal culpability or in good faith.” Star Scientific, 537 F.3d at 1366. This is one of those rare cases in which a finding of inequitable conduct is appropriate, particularly in light of the critical nature of the representations to the PTO in securing allowance of the '551 patent and the district court’s careful and thorough findings as to materiality and intent.

A Materiality

There is no question that the PTO rules, in particular Rule 56, require the submission of any known information that contradicts information submitted to the PTO:

Each individual associated with the filing and prosecution of a patent application has a duty of candor and good faith in dealing with the [PTO], which includes a duty to disclose to the [PTO] all information known to that individual to be material to patentability....
(b) Under this section, information is material to patentability when it is not cumulative to information already of record or being made of record in the application, and
(2) It refutes, or is inconsistent with, a position the applicant takes in:
*1301(i) Opposing an argument of unpatentability relied on by the [PTO], or
(ii) Asserting an argument of patentability.

37 C.F.R. § 1.56(a)-(b). These rules are directly relevant to the materiality component of inequitable conduct because “if a misstatement or omission is material under the ... Rule 56 standard, it is material [for purposes of inequitable conduct].” Digital Control, 437 F.3d at 1316; see also Pharmacia, 417 F.3d at 1373 (affirming a finding of inequitable conduct where a pri- or article by a declarant contradicted his declaration to the PTO). The district court found that statements made by Abbott’s predecessor to the EPO were highly material to the prosecution of the '551 patent because they contradicted representations Abbott made to the PTO regarding the membraneless sensor disclosed in the '382 patent. Trial Opinion, 565 F.Supp.2d at 1112-13. This finding is not clearly erroneous, and indeed is clearly correct.

1 PTO Representations

The original application that led to the '551 patent was filed in 1984. Over the next thirteen years, Abbott pursued the patent through half-a-dozen continuation applications that were repeatedly rejected for anticipation and obviousness, including repeated rejections over the '382 patent. In 1997, Abbott formulated a new argument for patentability during a brainstorming session attended by Lawrence Pope (“Pope”), Abbott’s patent attorney, and Dr. Sanghera, Abbott’s Director of Research and Development. Thereafter, new claims were drafted including electrochemical sensors lacking a membrane. During an interview with the examiner, Pope presented the new claims with the argument that the claims of the application were not anticipated or obvious because the claims taught a new glucose sensor that did not require a protective membrane when testing whole blood. Pope and the examiner specifically discussed the “[ojptionally, but preferably” language in the '382 patent. In Pope’s own summary of the interview, he stated that he “pointed out [to the examiner] that [the '382 patent] teaches that active electrodes designed for use with whole blood require a protective membrane.” J.A. 7644. The examiner then agreed that if Abbott produced “an affidavit or other evidentiary showing that at the time of the invention such a membrane was considered essential [for whole blood, it] would overcome [the '382 patent’s] teaching.” J.A. 7639.

In accordance with that agreement, Dr. Sanghera submitted a declaration to the PTO urging that the language of the '382 patent concerning the membraneless sensor should not be taken at face value. It stated in relevant part:

[B]ased on his historical knowledge he is confiednt [sic] that on the filing date of the earlist [sic] application leading to the present application on June 6, 1983 and for a considerable time thereafter one skilled in the art would have felt that an active electrode comprising an enzyme and a mediator would require a protective membrane if it were to be used with a whole blood sample. Therefore, he is sure that one skilled in the art would not read lines 68 to 65 of column J of U.S. Patent No. J, 5^5,882 to teach that the use of a protective membrane with a whole blood sample is optionally or merely preferred.

J.A. 7637 (emphases added). Pope, in submitting the affidavit,- further represented:

The art continued to believe [following the '882 patent] that a barrier layer for whole blood sample was necessary for a considerable period .... *1302One skilled in the art would not have read the disclosure of the Higgins patent (U.S. Jp,5J¡,5,382) as teaching that the use of a protective membrane with whole blood samples was optional. He would not, especially in view of the working examples, have read the “optionally, but preferably language at line 63 of column [f] as a technical teaching but rather mere patent phraseology. This is supported by the Declaration under 37 C.F.R. 1.132 of Gordon Sanghera which accompanies the present amendment.
... The applicants have established that a new claim limitation supported by the present application provides a patentable distinction over U.S. Patent No. 4,545,382, the key reference in the prosecution of the present application and its predecessors. There is no teaching or suggestion of unprotected active electrodes for use with whole blood specimens in this patent or the other prior art of record in this application.

J.A. 7645-46 (emphases added). Shortly thereafter, the PTO allowed the '551 patent with the newly added claims for a membraneless sensor.

2 EPO Representations

The district court found that Abbott had made directly contradictory representations to the EPO concerning the teaching of the '382 patent in European Patent EP 0 078 636 (“the '636 patent”)—a counterpart to the '382 patent with virtually identical specifications—and that Abbott had not disclosed those contradictory representations to the PTO.

In 1993, the '636 patent was revoked as obvious. One of the references the EPO relied on was a German reference labeled Dl. Abbott8 argued to the EPO that the Dl reference was distinguishable. On January 12, 1994, Abbott’s patent counsel submitted to the EPO a brief distinguishing the '636 patent from the Dl reference, stating:

Contrary to the semipermeable membrane of Dl, the protective membrane optionally utilized with the glucose sensor of the patent is [sic] suit is not controlling the permeability of the substrate (as set forth above under IV.2., in the membrane of Dl the permeability for the substrate must be kept on a low value to achieve a linear relationship between the measures [sic] currency and the substrate concentration in the test solution). Rather, in accordance with column 5, lines 30 to 33 of the patent in suit:
“Optionally, but preferably when being used on live blood, a protective membrane surrounds both the enzyme and the mediator layers, permeable to water and glucose molecules.”
See also claim 10 of the patent in suit as granted according to which the sensor electrode has an outermost protective membrane (11) permeable to water and glucose molecules. Finally, see Example 7 in column 10, lines 19 to 26 reporting that by using such a protective membrane the response time did not increase but from 24 to 60 sec. (without membrane) to 36-76 sec. (with membrane). Accordingly, the purpose of the protective membrane of the patent in suit, preferably to be used with in vivo measurements, is a safety measurement to prevent any course [sic] particles coming off during use but not a permeability control for the substrate.

J.A. 6530-31 (emphases added).

On May 23, 1995, Abbott’s patent counsel submitted another brief to the EPO *1303regarding the D1 reference and the '636 patent. It again discussed the “[o]ptionally, but preferably” language:

“Optionally, but preferably when being used on live blood, a protective membrane surrounds both the enzyme and the mediator layers, permeable to water and glucose molecules.”
It is submitted that this disclosure is unequivocally clear. The protective membrane is optional, however, it is preferred when used on live blood in order to prevent the larger constituents of the blood, in particular erythrocytes from interfering with the electrode sensor. Furthermore it is said, that said protective membrane should not prevent the glucose molecules from penetration, the membrane is “permeable” to glucose molecules.

J.A. 6585 (emphases added).

The district court found that the EPO statements contradicted the PTO representations in at least two significant ways. First, the district court found that by describing the “[optionally, but preferably” language as “unequivocally clear,” Abbott’s EPO representations contradicted Abbott’s representations to the PTO that a person having ordinary skill in the art would have understood the phrase as mere “patent phraseology” that did not convey a clear meaning. See Trial Opinion, 565 F.Supp.2d at 1110. Second, the district court noted that the EPO documents clearly explained that membranes were merely preferred for live blood. Id. at 1109-10. The documents identified reasons specifically directed toward live blood explaining why a membrane was necessary, supporting the position that these problems did not exist for in vitro testing of whole blood and that a membrane was not necessary for testing whole blood in vitro. See id. at 1109. These findings are not clearly erroneous, and indeed are manifestly correct.

3 Abbott’s Arguments

Abbott asserts that the district court erred in finding that the statements it made to the EPO were inconsistent with the PTO statements. It contends that in the EPO, the focus was on distinguishing the semipermeable membrane of the D1 reference from the protective, permeable membrane of the '636 patent. Abbott is correct that such a distinction was made in the EPO submissions. However, that is not the only point Abbott made in the EPO to distinguish the two references. The district court correctly found that Abbott also argued before the EPO that the protective membrane of the '636 patent was optional. The optional nature of the membrane was not irrelevant to the distinction of the D1 reference’s semipermeable membrane because the optional nature of the membrane proved that it was not the type of membrane required by the D1 reference, as Abbott appears to recognize.9 Abbott’s representation was particularly clear in the May 1995 submission. In labeling the “[o]ptionally, but preferably” language “unequivocally clear,” Abbott emphasized the alleged plain meaning of the language in the specification quoted in the immediately preceding sentence and reinforced that meaning in the following sentence, where it stated, “The protective membrane is optional, however, it is preferred when used on live blood in order to *1304prevent the larger constituents of the blood, in particular erythrocytes from interfering with the electrode sensor.” J.A. 6585. This sentence clearly was not directed to permeability. There can be no doubt that Abbott was making two assertions in this submission. It first explained that the protective membrane is optional but preferred when used on live blood and that the '382 patent language was “unequivocally clear” on this point. Then, using the transition “furthermore,” it expressed its second point about the permeability of the membrane.

The dissent suggests a somewhat different theory to explain Abbott’s representations to the PTO—a theory that significantly was not addressed by Abbott itself. Abbott’s theory is that the “unequivocally clear” statement of the second EPO submission was directed to the permeability of the membrane to water and glucose. As discussed above, this theory ignores the context of the “unequivocally clear” statement and particularly the sentence immediately following the “unequivocally clear” statement, which says nothing about permeability. The dissent attempts to fill this gaping hole in Abbott’s argument by suggesting that it would be reasonable to interpret the sentence following the “unequivocally clear” statement as also directed to how the membrane functioned, i.e., that it was “unequivocally clear” that it protected against fouling, rather than that it was “unequivocally clear” that it was optional. With respect, Abbott’s EPO statements cannot possibly be read in this manner.

The “unequivocally clear” statement is tied directly to the optional nature of the membrane when used with whole blood and other fluids. The three sentences in question read:

“Optionally, but preferably when being used on live blood, a protective membrane surrounds both the enzyme and the mediator layers, permeable to water and glucose molecules.”
It is submitted that this disclosure is unequivocally clear. The protective membrane is optional, however, it is preferred when used on live blood in order to prevent the larger constituents of the blood, in particular erythrocytes from interfering with the electrode sensor.

J.A. 6585. On its face, the description of the membrane’s protective function was itself designed to establish why the membrane was optional with whole blood and other fluids, but preferred with live blood. It was preferred with live blood because it protected against fouling. Both with respect to whole blood and live blood, it was “unequivocally clear” that the membrane was optional (whole blood and other fluids) or preferred (live blood). Necessarily, it was “unequivocally clear” that the membrane was not required for either whole blood or live blood. Thus, this is not a situation such as that referred to in Scanner Technologies Corp. v. ICOS Vision Systems Corp. in which “multiple reasonable inferences” as to the meaning of representations could be drawn from the evidence and there existed an “equally reasonable inference” favorable to Abbott. See 528 F.3d 1365, 1376 (Fed.Cir.2008).

Indeed, Abbott in oral argument agreed that the EPO statements were a “recharacterization of the sentence that was already before the [PTO] examiner.” Oral Argument at 40:40. Pope in his testimony agreed that the plain English reading of what Abbott told the EPO was contrary to what Abbott told the PTO.10 To *1305deprive an examiner of the EPO statements—statements directly contrary to Abbott’s representations to the PTO—on the grounds that they were not material would be to eviscerate the duty of disclosure. Moreover, if this could be regarded as a close case, which it is not, we have repeatedly emphasized that the duty of disclosure requires that the material in question be submitted to the examiner rather than withheld by the applicant. See LNP Eng’g Plastics, Inc. v. Miller Waste Mills, Inc., 275 F.3d 1347, 1361 (Fed.Cir.2001) (“[W]hen a question of materiality is close, a patent applicant should err on the side of disclosure.”); LaBounty Mfg., Inc. v. U.S. Int’l Trade Comm’n, 958 F.2d 1066, 1076 (Fed.Cir.1992) (“[A close case] makes it all the more necessary that the [reference] should [be] disclosed to the examiner. Close cases should be resolved by disclosure, not unilaterally by the applicant.”).

Abbott nonetheless contends that lawyer argument about prior art is not information material to patentability and that since both the EPO and the PTO representations were merely argument, any inconsistency between the two could not be material. This court has held that representations made to the PTO concerning the content of prior art “amount[] to mere attorney argument and our precedent has made clear that an applicant is free to advocate its interpretation of its claims and the teachings of the prior art.” Innogenetics, 512 F.3d at 1379. However, all of the cases Abbott cites involve patentees who simply made representations to the PTO about prior art in order to secure the allowance of their patents. See id., Young v. Lumenis, Inc., 492 F.3d 1336, 1344 (Fed.Cir.2007); Life Techs., Inc. v. Clontech Labs., Inc., 224 F.3d 1320, 1326 (Fed.Cir.2000); Akzo N.V. v. U.S. Int’l Trade Comm’n, 808 F.2d 1471, 1485 (Fed.Cir.1986). None of these cases involved a situation in which contradictory arguments made in another forum were withheld from the PTO. They do not speak to the applicant’s obligation to advise the PTO of contrary representations made in another forum. Before the EPO, Abbott made statements that contradicted the representations Abbott made to the PTO regarding the '382 patent. An applicant’s earlier statements about prior art, especially one’s own prior art, are material to the PTO when those statements directly contradict the applicant’s position regarding that pri- or art in the PTO. See 37 C.F.R. § 1.56(b)(2). In any event, the representations to the PTO were not merely lawyer argument; they were factual assertions as to the views of those skilled in the art, provided in affidavit form.

The district court’s finding that the EPO statements were highly material because they contradicted the position taken before the PTO was not clearly erroneous and was strongly supported by the uncontradicted record.

B Intent

In determining whether a failure to disclose material information was intentional, “the involved conduct, viewed in light of all the evidence, including evidence indicative of good faith, must indicate sufficient culpability to require a finding of intent to deceive.” Kingsdown, 863 F.2d at 876. Because direct evidence of deceptive intent is rarely available, such intent can be inferred from indirect and circumstantial evidence. Star Scientific, 537 F.3d at 1366.

*1306In concluding that Abbott’s representatives—Pope and Dr. Sanghera—intended to deceive the PTO by -withholding the EPO documents, the district court made five findings: (1) that the statements made to the PTO concerning the prior art '382 patent were absolutely critical in overcoming the examiner’s earlier rejections of the claims of the '551 patent; (2) that the EPO statements would have been very important to an examiner because they contradicted the representations made to the PTO; (3) that Pope and Dr. Sanghera both knew of the EPO statements and consciously withheld them from the PTO; (4) that neither Pope nor Dr. Sanghera provided a credible explanation for failing to submit the EPO documents to the PTO; and (5) that Pope’s and Dr. Sanghera’s explanations for withholding the EPO documents were so incredible that they suggested intent to deceive. See Trial Opinion, 565 F.Supp.2d at 1113-16. The first of these is undisputed. The second is not clearly erroneous as discussed above. The third is also undisputed; Pope and Dr. Sanghera together evaluated the documents from the EPO proceedings during the prosecution of the '551 patent and made a conscious decision to withhold them from the PTO.

The district court’s fourth and fifth findings regarding intent are based on the district court’s assessment of witness credibility. “[Djetermination of the credibility of the witnesses is within the discretion of the presiding official who heard their testimony and saw their demeanor,” Griessenauer v. Dep’t of Energy, 754 F.2d 361, 364 (Fed.Cir.1985), such that a judge’s credibility determinations are “virtually unreviewable,” Hambsch v. Dep’t of the Treasury, 796 F.2d 430, 436 (Fed.Cir.1986). As such, “[t]his court may not reassess, and indeed is incapable of reassessing, witness credibility and motive issues on review.” LNP Eng’g, 275 F.3d at 1361.

The district court judge had the opportunity to observe both Pope’s and Dr. Sanghera’s testimony during the bench trial and found their explanations neither plausible nor credible. In fact, the court explicitly noted that “Attorney Pope did not prove to be a convincing trial witness.” Trial Opinion, 565 F.Supp.2d at 1113. The district court similarly found Dr. Sanghera’s testimony not to be credible, expressing twice in its opinion that his demeanor was unconvincing. See id. at 1115 (“As a trial witness, it must be said that Dr. Sanghera was impeached on substantive points with his prior inconsistent statements and exhibited an unconvincing demeanor.” (citation omitted)); id. at 1116 (“His unconvincing trial demeanor has been a factor in this determination.”).

These findings are amply supported. One example will suffice. In attempting to dismiss the EPO statements as merely cumulative, Pope was confronted with the need to explain the May 1995 statement that the “[ojptionally, but preferably” language was “unequivocally clear.” Pope testified that “it was my understanding at the time that what was unequivocally clear was that that the protective membrane, which was sometimes required by the '636 patent was permeable to water and glucose.” J.A. 2983. This is contrary to the explicit meaning of the May 1995 statement, as the district court found. When the district court asked Pope to further explain the sentence, “It is submitted that this disclosure is unequivocally clear,” Pope responded:

I would agree with Your Honor that as a matter of normal English construction, the sentence, “it is submitted” refers to what comes immediately before and refers to all of it, but as a patent attorney reading this document and making a judgment call, I was making *1307my judgment call on what was being conveyed, what information was important, what position was being taken. And in that context, I didn’t believe that they were trying to convey anything about the optional but preferably language.

J.A. 2990. The district court did not err in finding that Pope’s “trial explanation for his withholding [the EPO submissions] was not plausible and he was not credible” and that Pope “acted with specific intent to deceive [the examiner] and the PTO,” Trial Opinion, 565 F.Supp.2d at 1113, particularly in view of Pope’s admission that the EPO statement’s “normal English construction” directly contradicted his representations to the PTO on a critical issue.

So too Dr. Sanghera testified that he considered it appropriate to withhold from the PTO his own knowledge, reflected in the EPO submission, that “general English usage” contradicted the representations that he had made to the PTO. He testified that “in general English usage, [he] would not use the terms ‘optional’ or ‘preferable’ to describe something that is required,” and he could not recall “any instance during the course of [his] scientific career in which [he] use[d] the terms ‘optional’ or ‘preferable’ to refer to something that was required.” J.A. 3009. Yet like Pope, Dr. Sanghera testified that representations to the EPO confirming that the language was “unequivocally clear” could be withheld. The district court did not err in finding bad faith.

The district court’s conclusion of intent also finds support in the fact that Pope relied on Dr. Sanghera, who was not a person having ordinary skill in the art at the time of the '382 patent, to provide a declaration to the PTO regarding the teachings of the '382 patent rather than the inventors of the patent, who appear to have had a view quite contrary to Abbott’s opinion as to the language of the '382 patent. Dr. Higgins, an inventor on both the '382 and the '551 patents, testified that even the phrase “preferably when being used on live blood” meant that “you would probably need a membrane, but not definitely.” J.A. 3104. He felt that Dr. Sanghera’s declaration statements concerning whole blood were simply “wrong.” J.A. 3708. “Had [Dr. Sanghera] been around in the field, working in the field in 1983, and he had seen our patent specification, I believe he, as a person skilled in the art, would have concluded that the membrane may not be necessary.” J.A. 3707-OS.

Finally, Abbott argues that Dr. Sanghera demonstrated a lack of intent to deceive by satisfying his duty to disclose by providing the EPO documents to Pope. See 37 C.F.R. § 1.56(d) (“Individuals other than the attorney, agent or inventor may comply with this section by disclosing information to the attorney, agent or inventor.”). However, the district court was correct to note that by submitting a declaration to the PTO, Dr. Sanghera was still obligated to avoid intentional deception. See Trial Opinion, 565 F.Supp.2d at 1115. “Cases involving affidavits or declarations are held to a higher standard.” Innogenetics, 512 F.3d at 1379. This court has in the past expected more of declarants before the PTO. In Ferring B.V. v. Barr Laboratories, Inc., we held that the failure to disclose possible biases of the declarants constituted a material omission where the declarations were submitted to overcome a prior art rejection. 437 F.3d 1181, 1190 (Fed.Cir.2006). Refac International, Ltd. v. Lotus Development Corp. involved similar circumstances in which declarations submitted to establish enablement failed to disclose possible biases. 81 F.3d 1576, 1581 (Fed.Cir.1996). In this case, it was Dr. Sanghera’s duty to avoid intentional *1308deception in his declaration before the PTO, and merely disclosing the EPO documents to Pope did not obviate that duty.

The dissent sets up various straw men that are irrelevant to the district court’s finding of intent and the majority’s affirmance of the intent issue. Thus, the dissent urges that Pope and Dr. Sanghera were not aware of U.S. Patent No. 4,388,-166 to Suzuki et al.—a prior art reference disclosing advantages and disadvantages of using membraneless sensors with whole blood—and were not aware of Dr. Higgins’s contrary interpretation of the “[o]ptionally, but preferably” language. This is undisputed; the district court did not base its finding on intent on the existence of such knowledge. Nor did the district court find that Abbott misrepresented Dr. Sanghera’s qualifications before the PTO.

More to the point is the dissent’s urging that Pope’s and Dr. Sanghera’s testimonies reflect a plausible conclusion that the EPO submissions were not material. For the reasons described above, the district court was clearly correct in concluding that the explanations offered by Pope and Dr. Sanghera were not plausible. The dissent’s efforts to come up with new theories of plausibility (discussed above in connection with materiality) are both unconvincing and beside the point. Pope’s and Dr. Sanghera’s intent must be judged by their asserted beliefs at the time, not by some imputed theory that Abbott itself does not even offer on appeal.

Accordingly, the district court did not clearly err in finding that Pope and Dr. Sanghera both intended to deceive the PTO by withholding the EPO documents.

Because the district court’s findings that the EPO submissions were highly material to the prosecution of the '551 patent and that Pope and Dr. Sanghera intended to deceive the PTO by withholding those submissions were not clearly erroneous, the district court did not abuse its discretion in holding the '551 patent unenforceable due to inequitable conduct.

Ill Noninfringement of the '164 and '745 Patents

On summary judgment, the district court concluded that there was no issue of material fact as to whether the BD Test Strip infringes the '164 and '745 patents because the BD Test Strip does not satisfy the “non-flowing manner” limitation present in both patents. Abbott contends that the district court’s decision is based on an erroneous claim construction. We review both the district court’s grant of summary judgment and its claim construction de novo. AquaTex Indus., Inc. v. Techniche Solutions, 479 F.3d 1320, 1328 (Fed.Cir.2007); Cybor Corp. v. FAS Techs., Inc., 138 F.3d 1448, 1456 (Fed.Cir.1998) (en banc).

The '164 patent describes a glucose sensor for blood samples smaller than one microliter. The '745 patent describes a glucose sensor for blood samples smaller than one microliter that employs diffusible mediators. A limitation in each of Abbott’s asserted claims in the '164 and '745 patents requires that the blood sample being analyzed be held within the strip in a “non-flowing manner.” Claim 16 of the '164 patent is representative in this respect:

16. A method for determining a concentration of an analyte [such as glucose] in a body fluid of a patient, comprising the steps of:
creating an unassisted flow of a body fluid from the patient;
transporting a portion of the body fluid into an analyte sensor configured and arranged to determine the *1309concentration of the analyte from 500nL or less of body fluid;
holding the sample in a non-flowing manner within a sample chamber of the analyte sensor; and
determining the concentration of the analyte in the body fluid from the portion of the body fluid transported into the analyte sensor.

'164 patent col.25 1.66-col.26 1.11 (emphasis added). The district court construed the claim term “holding the sample in a non-flowing manner within the sample chamber of the analyte sensor” to mean that “the sample is not moving in the sample chamber during the measurement.” Summary Judgment Order, 560 F.Supp.2d at 854. Even a sample “flowing at a very slow rate” is “directly at odds with the plain and ordinary meaning of ‘non-flowing,’ ” the court stated. Id.

At summary judgment, the district court applied its construction, concluding that the limitation is not satisfied even if the only motion present in the sample is convective flow. See id. at 856. Abbott argues that such an expansive reading is unwarranted because all liquids have at least a minor amount of convective flow, pointing to the uncontroverted testimony of its expert that “[t]here is always ... at least a minor amount of convection occurring within any liquid.” J.A. 10553. We agree with Abbott. Here any reasonable construction must allow for certain types of motion that are inherent within liquids, including convective flow and Brownian motion.11

Even though we reject the district court’s broad reading of the “non-flowing manner” limitation, we also reject Abbott’s alternative construction. Abbott contends that “flow” means to move in a stream and that the proper construction of the “non-flowing manner” limitation should be that the sample is not moving through the sample chamber during measurement. Abbott analogizes to a river or stream that stops flowing when dammed. However, the plain meaning of “flow” is not so limited. A dammed river or stream may no longer flow along the same path as it once did, but the water will continue to flow within the lake behind the dam—a flow that is more than mere convective flow. The sample chamber on the BD Test Strip is no different. Blood can flow through the sample chamber or it can flow within the sample chamber. In either case the blood is still flowing. We construe a “non-flowing manner” to mean that a sample “is not moving in the sample chamber during the measurement,” other than motion attributable to convective flow, Brownian motion, or other qualities inherent to all liquids.

Under the correct claim construction, summary judgment of noninfringement was still appropriate. Abbott on appeal does not contend that the BD Test Strip does not evidence flow under the claim construction that we have adopted, and indeed Abbott’s expert admitted that there is a “slight swirling in the chamber.” J.A. 12258. BD/Nova presented uncontradicted evidence that the BD Test Strip does not infringe Abbott’s '164 and '745 patents because the blood sample flows within the strip’s measurement zone during the entire measurement period. Expert testimony and numerous video recordings showed that blood samples on the BD Test Strip exhibit a swirling motion. The evidence also showed that the swirling motion persists during the entire five-second measuring time of the BD Test Strip. Therefore, on this record, summary judgment of noninfringement of the '164 and *1310'745 patents by the BD Test Strip was proper.

IV Anticipation of '745 Patent Claims

On summary judgment, the district court also concluded that there was no material issue of fact as to the anticipation of claims 1-5, 8, 21-28, 28, 31, and 34 of the '745 patent because PCT application WO 98/35225 (“the1 '225 reference”) already disclosed diffusible mediators. Summary Judgment Order, 560 F.Supp.2d at 875. “This court reviews a ruling of summary judgment de novo. Although anticipation under 35 U.S.C. § 102 is a question of fact, it may be decided on summary judgment if the record reveals no genuine dispute of material fact.” Golden Bridge Tech., Inc. v. Nokia, Inc., 527 F.3d 1318, 1321 (Fed.Cir.2008) (citation omitted).

The '225 reference contains the same disclosure as the '164 patent. Summary Judgment Order, 560 F.Supp.2d at 874. It describes a glucose sensor for blood samples smaller than one microliter. The primary improvement of the '745 patent over the '225 reference and the '164 patent is the use of diffusible or leachable mediators with such sensors. Claim 1 of the '745 patent reads:

1. A method for determining a concentration of glucose in a sample, comprising the steps of:
(a) contacting a sample with an electrochemical sensor comprising:
(i) an electrode pair comprising a working electrode and a counter electrode, wherein the working electrode and counter electrode are separated by a closest distance in a range of 200 to 1000 ixm;
(ii) a measurement zone positioned adjacent to the working electrode and the counter electrode, wherein the measurement zone is sized to contain a volume of no more than about 1 jxL of the sample; and
(iii) an analyte-responsive enzyme and a diffusible redox mediator disposed in the measurement zone;
(b) holding the sample within the measurement zone in a non-flowing manner;
(c) generating a sensor signal at the working electrode within a measurement period of no greater than about 5 minutes, wherein a background signal that is generated by the redox mediator is no more than five times a signal generated by oxidation or reduction of 5 mM of glucose-, and
(d) determining the concentration of the glucose using the sensor signal.

'745 patent col.61 11.39-63 (emphases added).

On appeal, Abbott challenges only whether the '225 reference disclosed the “diffusible redox mediator” limitation and the “background signal” limitation. However, Abbott did not argue the “background signal” issue to the district court in its opposition to summary judgment. See Summary Judgment Order, 560 F.Supp.2d at 871-72. Therefore, Abbott can not do so now on appeal and has waived the issue. See Fresenius USA Inc. v. Baxter Int’l, Inc., 582 F.3d 1288, 1296 (Fed.Cir.2009) (“If a party fails to raise an argument before the trial court, or presents only a skeletal or undeveloped argument to the trial court, we may deem that argument waived on appeal.... ”).

Thus, the only remaining issue is the “diffusible redox mediator” element. The district court concluded that there was *1311no genuine issue of material fact on whether the '225 reference disclosed the “diffusible redox mediator” element of the '745 patent because the '225 reference “acknowledge[s] the possibility of using a leaching or diffusing mediator, although clearly the drafters preferred those that did not leach or diffuse.” Summary Judgment Order, 560 F.Supp.2d at 873. Abbott argues that the district court focused on a single passage out of context. However, the plain language of the '225 reference does disclose diffusible or leachable mediators.

The '225 reference states:
Preferably, there is little or no leaching of the redox mediator away from the working electrode 22 into the sample during the measurement period, which is typically less than about 5 minutes. More preferably, the redox mediators of the present invention are bound or otherwise immobilized on the working electrode 22 to prevent undesirable leaching of the mediator into the sample. A diffusing or leachable (i.e., releasable) redox mediator is not desirable when the working and counter electrodes are close together....

'225 reference at 9. The plain meaning of the passage implies that diffusible mediators are contemplated, though they are not preferred. This conclusion is reinforced by the claims of the '225 reference. Claim 127 of the reference describes a method calling for a “redox mediator.” It does not specify whether this mediator is a leachable or a non-leachable mediator—that is, diffusible or non-diffusible. Claim 128, however, encompasses “[t]he method of claim 127, wherein the redox mediator is a non-leachable redox mediator,” i.e., a nondiffusible mediator. '225 reference at 63. “[T]he presence of a dependent claim that adds a particular limitation gives rise to a presumption that the limitation in question is not present in the independent claim.” Phillips v. AWH Corp., 415 F.3d 1303, 1315 (Fed.Cir.2005) (en banc). Here, where the dependent claim adds only the non-leachable (non-diffusible) mediator limitation, this principle of claim construction compels the conclusion that the '225 reference claims diffusible as well as nondiffusible mediators in claim 127. As discussed above, the claims here must be considered part of the disclosure.

Moreover, not a single expert in the case challenged the fact that the '225 reference disclosed diffusible mediators, either explicitly or implicitly. While the witnesses argued that the '225 reference at least to some extent taught away from their use, “[a] reference is no less anticipatory if, after disclosing the invention, the reference then disparages it.... [T]he question whether a reference ‘teaches away’ from the invention is inapplicable to an anticipation analysis.” Celeritas Techs., Ltd. v. Rockwell Int’l Corp., 150 F.3d 1354, 1361 (Fed.Cir.1998). Therefore, there was no triable issue of material fact left for a jury to consider on the issue of whether the '225 reference disclosed diffusible mediators. As such, summary judgment on anticipation of the '745 patent was proper.

CONCLUSION

For the aforementioned reasons, we affirm the district court’s judgment that claims 1-4 of the '551 patent are invalid due to obviousness and the entire patent is unenforceable because of inequitable conduct. We also affirm the district court’s grant of summary judgment of noninfringement of the '164 and '745 patents and invalidity of claims 1-5, 8, 21-23, 28, 31, and 34 of the '745 patent due to anticipation.

AFFIRMED

*1312COSTS

No costs.

. Because Abbott accuses BD and Nova of infringement based on the same BD products, BD and Nova will hereinafter be referred to collectively as “BD/Nova.”

. Abbott also sued BD/Nova alleging that the BD Test Strip infringes U.S. Patent No. 5,628,890. That patent was the subject of a separate jury trial in which the jury found that the BD Test Strip did infringe the patent, but that the patent was invalid. After entry of judgment, Abbott appealed, and that judgment is the subject of a separate appeal in Therasense, Inc. v. Becton, Dickinson & Co., Nos.2009-1008, -1009, -1010,-1034, -1035, - 1036, -1037, 593 F.3d 1325, 2010 WL 254904 (Fed.Cir. Jan. 25, 2010).

. Claim 1 of the '551 patent is representative. It provides:

1. A single use disposable electrode strip for attachment to the signal readout circuitry of a sensor to detect a current representative of the concentration of a compound in a drop of a whole blood sample comprising:
a) an elongated support having a substantially flat, planar surface, adapted for releasable attachment to said readout circuitiy;
b) a first conductor extending along said surface and comprising a conductive element for connection to said readout circuitry;
c) an active electrode on said strip in electrical contact with said first conductor and positioned to contact said whole blood sample;
d) a second conductor extending along said surface comprising a conductive element for connection to said read out circuitry; and
e) a reference counterelectrode in electrical contact with said second conductor and positioned to contact said whole blood sample, wherein said active electrode is configured to be exposed to said whole blood sample without an intervening membrane or other whole blood filtering member and is formed by coating a portion of the first conductor with a mixture of or layers of an enzyme which catalyzes a redox reaction with said compound in whole blood and a mediator compound which transfers electrons from said redox reaction to said first conductor to create a current representative of the concentration of said compound in said whole blood sample; and wherein said active electrode which is formed on a portion of said conductor is not in electrical contact with said reference counterelectrode but these electrodes are so dimensioned and positioned that they can be simultaneously completely covered by a single drop of whole blood such that this drop provides an electrical path between these electrodes to support said current representative of the concentration of said compound in said whole blood sample.

'551 patent col. 13 1.29-col. 14 1.17 (emphasis added).

. However, one of the inventors of the '382 and '551 patents, Dr. Higgins, testified that even the phrase "preferably when being used on live blood" meant that "you would probably need a membrane, but not definitely.” J.A. 3104.

. Abbott also attempts to rely on U.S. Patent No. 4,897,173 ("the '173 patent”) to show that the '382 patent did not disclose a membraneless sensor for whole blood and that the conventional wisdom of those of ordinary skill in the art was that a membrane was necessary when testing whole blood even after the '382 patent. The fact that the '173 patent contained an example in which whole blood was tested using a sensor including a membrane hardly suggests that a membrane was universally required.

. The '551 invention's oxygen sensitivity, characterized as “minimal,” is only slightly better at 4%. See '551 patent col.7 11.18-20.

. Abbott also relies on two other alleged differences between the '382 and '551 patents: (1) the use of screen printing to make the '551 electrode, and (2) the use of oxidized electrodes on the '382 patent as opposed to the '551 patent's nonoxidized electrodes. However, these distinctions do not relate to any claimed elements in the '551 patent, and they can hardly be relied upon to show the '551 patent’s nonobviousness.

. At this time, the party operating before the EPO was Medisense, a predecessor to both Therasense and Abbott. We use the term “Abbott” to include Medisense.

. In describing the distinction made before the EPO between the '636 patent and the D1 reference, Abbott states in its reply brief, “If that kind of [semipermeable] membrane is needed, it is needed whenever glucose is measured, regardless of the liquid—buffer, blood, or interstitial fluid. Thus, by showing that the '382/'636 did not use a membrane for at least some liquids, [Abbott] proved that it did not use the Dl's glucose-controlling membrane.” Pl.-Appellants' Reply Br. 13 (citation omitted).

. Pope agreed that "as a matter of normal English construction," the "unequivocally clear” statement "refers to [the '[ojptionally, but preferably’ language] and refers to all of *1305it.” J.A. 2990. Thus, according to Pope, "if [the ‘[ojptionally, but preferably'] language stood alone, ... it would definitely indicate that the use of a protective membrane was optional or preferable.” J.A. 2988.

. The district court did not decide whether Brownian motion would be covered by the claim construction. Summary Judgment Order, 560 F.Supp.2d at 856.

. These were the only two ways in which Abbott’s submissions were found to be "inconsistent” and thus material under 37 C.F.R. § 1.56(b)(2). Thus, it is these two findings that must be reviewed for clear error. Although the district court cited the "reasonable examiner” standard and found that Abbott's EPO submissions would have been "highly material” to such an examiner, the district court made this finding only to the extent that the submissions were "contradictory” or inconsistent. Trial Opinion, 565 F.Supp.2d at 1110.