Danisco A/S v. Novozymes A/S , 427 F. Supp. 2d 443 ( 2006 )

427 F. Supp. 2d 443 (2006)

DANISCO A/S and Danisco USA Inc., Plaintiffs-Counterdefendants,
v.
NOVOZYMES A/S and Novozymes North America, Defendants-Counterclaimants.

No. 05 Civ.1972(GEL).

United States District Court, S.D. New York.

April 17, 2006.

Michael F. Borun and Kevin M. Flowers, Marshall, Gerstein & Borun LLP, Chicago, IL, for Plaintiffs-Counterdefendants.

Joseph R. Robinson, Steven E. Lipman, and Kevin L. Reiner, Darby & Darby P.C., New York, NY, for Defendants-Counterclaimants.

Steven M. Amundson, Frommer Lawrence & Haug LLP, New York, NY, for Frommer Lawrence & Haug LLP.

OPINION AND ORDER

LYNCH, District Judge.

In this patent infringement action, defendant Novozymes moves to compel production of certain documents withheld by plaintiff Danisco on grounds of attorney-client or work product privilege, arguing that these privileges are inapplicable because the documents were generated in furtherance of a fraud on the patent office. The motion will be denied.

For purposes of this motion, Novozymes does not dispute that the withheld documents would ordinarily come within the protection of the attorney-client or work product privilege. However, communications otherwise protected by these privileges lose protection if they relate to client communications made in furtherance of criminal or fraudulent conduct. See In re Grand Jury Subpoena Duces Tecum, 731 F.2d 1032, 1038 (2d Cir.1984) ("It is well-established that communications that otherwise would be protected by the attorney-client privilege or the attorney work' product privilege are not protected if they relate to client communications in furtherance of contemplated or ongoing criminal or fraudulent conduct."). The party invoking this exception must show probable cause to believe that a crime or fraud has been attempted and that the communications in question were made in furtherance of that fraud. See In re Richard Roe, Inc. (Roe II), 168 F.3d 69, 70 (2d Cir.1999). "This standard has been rephrased as requiring `that a prudent person have a reasonable basis to suspect the perpetration of a crime or fraud, and that the communications were in furtherance thereof.'" In re John Doe, Inc., 13 F.3d 633, 634 (1994), quoting In re Grand Jury Subpoena Duces Tecum, 731 F.2d 1032, 1039 (2d Cir.1984). As the Second Circuit has stressed, to properly override the privilege, a court must determine whether each communication at issue was made in furtherance of a crime or fraud. Id. at 71; In re Richard Roe, Inc. (Roe I), 68 F.3d 38, 40 (2d Cir.1995).

Almost twenty years ago, the Federal Circuit noted that claims of inequitable conduct and fraud on the Patent Office have become "an absolute plague." Burlington Indus., Inc. v. Dayco Corp., 849 F.2d 1418, 1422 (Fed.Cir.1988). Such charges are easily made, and it is also easy to cobble together the rule that only probable cause to believe a fraud has been committed is necessary for the crime-fraud exception to apply, with the rule that a patent applicant has a strong "duty of candor and good faith in dealing with the [Patent and Trademark] Office," 37 C.F.R. § 1.56(a), to turn a good faith scientific dispute or a minor error of fact in a patent application into a specious argument that attorney-client privileges should be invaded because a fraud has been committed. However, the attorney-client privilege serves important values, Swidler & Berlin v. United States, 524 U.S. 399, 403, 118 S. Ct. 2081, 141 L. Ed. 2d 379 (1998) (observing that the "attorney-client privilege is one of the oldest recognized privileges for confidential communications" and "is intended to encourage full and frank communication between attorneys and their clients and thereby promote broader public interests in the observance of law and the administration of justice" (quotation marks omitted)), and the rule permitting that privilege to be invaded when it is abused for purposes of engaging in fraud is aimed at serious misconduct, In re Grand Jury Investigation, 399 F.3d 527, 535 (2d Cir.2005) (noting that rules such as the crime-fraud exception have been developed "to limit egregious abuses of the protections that the privilege affords" (emphasis added)). While courts must be vigilant to prevent a client's attempt to misuse attorneys to perpetrate a crime or a genuine fraud, the privilege is not to be lightly cast aside based on speculative allegations. Cf. Roe II, 168 F.3d at 71 (noting exceptions to privilege, including crime-fraud exception, "should not be framed so broadly as to vitiate much of the protection they afford").

Despite the lengthy briefs and voluminous supporting documents submitted by both sides on the applicability of the crime-fraud exception, the issues presented are relatively simple. Danisco seeks to enforce a patent concerning methods of baking and preparing dough by adding an enzyme to the dough that breaks down certain fat molecules in the dough, called lipids ("the '346 Patent"). In seeking the patent, Danisco claimed that its enzyme ("Lipase 3") was valuable and original because it worked to break down three distinct types of lipids: triglycerides, glycolipids (in particular, a lipid known as "DGDG"), and phospholipids. (D. Mem. 1, 10; P. Mem. 2-3.) The ability to hydrolyze all three types of lipids with a single enzyme was claimed to represent a significant advance on prior art. (P. Mem.6.) Novozymes claims that Danisco defrauded the Patent Office in three ways: (1) by withholding information that its enzyme in fact did not work against phospholipids (D.Mem.32-33); (2) by misrepresenting to the Patent Office that an already-extant enzyme manufactured by Danisco did not work against glycolipids (id. at 34-35); and (3) by misrepresenting the content or significance of certain Novozymes patent documents that could be construed as prior art (id. at 36-38).

Needless to say, the chemical transformations involved are highly sophisticated, and no doubt reasonable scientists can disagree about the effectiveness of various enzymes, or about the interpretation of the scientific evidence developed through various tests on the products in question. The Court does not purport to resolve these scientific disputes, or to reach any conclusions on the underlying merits of the patent case, on the basis of the record presented on a discovery motion. Upon careful examination of the materials submitted, however, the Court is firmly persuaded that Novozymes had failed to establish any reasonable basis for a belief that anyone involved in prosecuting Danisco's patent intended to deceive or to commit fraud.

1. Does Danisco's Enzyme Work Against Phospholipids?

Novozymes asserts that Danisco obtained a patent for an enzyme that hydrolyzes, among other things, phospholipids, when in fact the enzyme does not work as claimed, and Danisco knew that it did not. This dramatic accusation rests on a single statement in a separate international patent application which Novozymes claims demonstrates that one of the inventors listed on the patent in suit knew that the enzyme did not work against phospholipids. (D. Mem. 32, citing WO 01/39602, Novozymes Ex. 27, at 2.)

The scientific evidence, to the extent the Court can decipher it, may be subject to different interpretations. Danisco points out that in the very application relied on by Novozymes, a table reporting the results of testing "shows unequivocally that Lipase 3 has a phospholipase activity [i.e., activity against phospholipids] of 24 PLU/g." (P. Mem.18.) Moreover, a table in the '346 Patent itself shows a general decrease in the presence of a particular phospholipid (phosphatidyl choline) when Lipase 3 is added to dough. (P. Mem. at 18, citing Novozymes Ex. 2 clmn. 31, tbl. 8.4.) These results may well be challenged. For example, the table in the '346 Patent actually shows an increase in phosphatidyl choline when 5000 LUS/kg of Lipase 3 is present, followed by a significant decrease as the amount of Lipase 3 increases. (Id.) And Novozymes points to another table in the '346 Patent allegedly showing ineffectiveness against phosphatidyl choline. (Reply 9, citing Novozymes Ex. 2 clmn. 32, tbl. 8.7.) Moreover, the study in the WO 01/39602 application relied on by Danisco may be questionable because the test was performed on lecithin powder in water rather than on dough, or because "phospholipase activity of 24 PLU/g" may not be all that terrific.^[1] (Reply 8-9.) However, there is no allegation whatsoever that the test results submitted to the Patent Office were faked or inaccurate, and any weakness in the data on phosphatidyl choline, or inconsistency of those data with other tests regarding phospholipids, is apparent on the face of the data and was available to the patent examiner to permit the examiner to draw his or her own conclusions.

Assuming arguendo that a reasonable scientist could believe that the case for the phospholipase activity of Lipase 3 was not made, to make out probable cause for a finding of fraud Novozymes needs to present evidence that Danisco and its representatives did not subjectively believe that Danisco's product worked. The sole Danisco statement on which Novozymes relies in this regard is a statement in an international patent application by one of the inventors of Lipase 3 (Dr. Jrn Se) that Novozymes summarizes as asserting "that the same lipase disclosed in the Danisco '394 patent application [relating to the '346 Patent] `did not have a phospholipid hydrolysing [sic] effect.'" (D. Mem. 32, quoting Novozymes Ex. 27, at 2.)^[2] But the reference in Dr. Se's prior statement does not clearly refer to Lipase 3. What Dr. Se actually says is that studies had shown that "the Aspergillus niger lipase did not have a phospholipid hydrolysing [sic] effect." (P. Mem. at 19, quoting Novozymes Ex. 27, at 2.) The parties dispute at some length what this means. Danisco maintains that Lipase 3 is derived from Aspergillus tubigensis, which it claims is a different fungus than Apergillus niger, and that the unsuccessful Aspergillus niger enzyme referred to in the quoted extract must be one of the "prior art lipases" that Dr. Se discusses elsewhere in the international application referred to. (Id. at 19-20.) Novozymes retorts that Apergillus tubigensis is merely a variant of Aspergillus niger, so that Lipase 3 could plausibly be referred to as an "Aspergillus niger lipase," and that the crucial quotation refers to the singular "lipase disclosed in" a particular prior application (apparently Lipase 3), whereas the prior art references cited by Danisco refer to a number of different lipases. (Reply 10.)

Novozymes may have the better of the argument as to the proper interpretation of this statement. But even if Novozymes's interpretation is accepted, a disputed and cryptic "admission" of this sort is a slim reed upon which to hang an accusation of fraud. Danisco has consistently sought patents all over the world for Lipase 3, touting its virtues the while. That one of its scientists may have at some point suggested, in a different context, without supporting data, that the enzyme has not been shown to hydrolyze phospholipids does not render Danisco guilty of fraud when it subsequently claimed, in connection with the '346 Patent application, with supporting data available for inspection, that it did.^[3]

2. Does Danisco's Prior Enzyme Work against Glycolipids?

Novozymes next charges that Danisco fraudulently overstated the advance represented by Lipase 3 over a prior Danisco product ("Exel"), by asserting that the lipase in that product did not break down glycolipids, when the evidence showed that it did. Ironically, in this case Novozymes and Danisco change places: with respect to the phospholipase action of Lipase 3, Danisco seizes on arguably minor evidence of the enzyme's effectiveness to defend its claims, while Novozymes derides the same evidence as tantamount to nothing at all; with respect to the effectiveness of Exel against glycolipids, Novozymes makes the same kinds of arguments from weak evidence that it characterizes as fraudulent in the former dispute, while Danisco takes up the cause of pooh-poohing weak evidence. The briefs thus inadvertently demonstrate that the scientific arguments are disputable, but the claims of "fraud" are mostly lawyers' spin.

The critical assertion in the Lipase 3 patent application is the claim, made in direct response to the patent examiner's rejection of the claim that Lipase 3 did not represent an advance over Exel, that Exel (unlike Lipase 3) has "no activity towards glycolipids." (Novozymes Ex. 29, at 10-11; see also id. Ex. 30, at 2-3.) Novozymes vigorously charges that this representation was fraudulent, because Danisco was aware of test results showing that Exel did indeed have some success in hydrolyzing certain glycolipids.

The centerpiece of this argument is a study known to Danisco's Dr. See that could be read to show that the lipase in Exel "hydrolyzes up to 3% of the glycolipid DGDG." (D.Mem.15-16.) Once again, neither party does very much to document to a non-chemist whether this is or is not impressive glycolipase activity. However, Dr. See has consistently cited this very study to mean that the "effect of the two commercially available lipases [including the Lipase in Exel] are [sic] negligible." (P. Mem. 23, quoting Novozymes Exs. 24 at 26, 25 at 62.) Novozymes responds not by disputing Dr. Soe's interpretation of the data, but by triumphantly declaring this proof of fraud, since "`negligible' does not mean `none.'" (Reply 11.) Indeed it does not, and if anything turns on whether Exel has any glycolipase capability, a fact-finder may well conclude, when all the scientific evidence, including qualified expert testimony, is properly before it, that it does. It may even turn out, upon examination, that such activity is more than negligible. But when the issue is whether a new product that allegedly has significant ability to hydrolyze glycolipids represents a patentable advance on another product already on the market, the difference between "no" and "negligible" is not so significant that, essentially without more, the Court can infer an intent to defraud from the use of the former by a party which had in another context used the latter. This is particularly so given that Danisco maintains that numerous other experiments conducted by Dr. Søe and others showed that the lipase in Exel had no activity against glycolipids such as DGDG and MGDG (P. Mem.24-27), test results that Novozymes does not address, let alone dispute, in its reply submission. (Reply 11-13.)

Danisco also maintains that the data cited by Novozymes showing that the lipase in Exel had activity against glycolipids are unreliable in any case (and were known to be so when the '346 Patent application was made). As with the question referred to above concerning the relevance of a study performed in a medium other than dough, there appears to be a legitimate scientific dispute about the reliability of Gas-Liquid Chromatography in measuring glycolipids in dough. (P. Mem. 28-29; Reply 13.) The Court takes no position on this controversy. But the existence of the dispute supports the inference that Danisco's conclusions about the ineffectiveness of Exel against glycolipids is not rendered manifestly incorrect by the data cited by Novozymes. Similarly, Danisco cites evidence that Novozymes's own scientists also took the position that enzymes like Exel were not active against DGDG. (P. Mem. 27-28, citing Izraelewicz Decl. Ex. G, at 16.) Novozymes is quite correct that what matters is not what its team thought, but whether Danisco's agents believed what they told the Patent Office. (Reply 13.) But in the absence of any direct evidence of Danisco's subjective beliefs about Exel—and there is none—the inference of fraudulent intent can only be drawn from evidence that what Danisco said could not reasonably be believed in good faith. If Novozymes thought the same thing, it is hard put to argue that anyone else taking that position must be a liar.

3. Did Danisco Misrepresent Novozymes's Lipopan F Patents?

Novozymes's final claim of fraud is more complicated to explain. In support of the Lipase 3 patent, Danisco argued to the patent office that the enzyme disclosed in certain earlier Novozymes patent documents did not have the critical quality of effectiveness against all three lipids: triglycerides, glycolipids, and phospholipids. (D.Mem.36.) Principally, Novozymes claims that the product disclosed in those documents was actually Lipopan F, the Novozymes product which Danisco asserts in this action infringes the '346 Patent, and that Danisco knew this fact but failed to mention it to the Patent Office. The patent examiner ultimately concluded that the earlier Novozymes product to which these patents applied ("Lipopan F") did have the claimed triple effect when used in dough products, but allowed the patent because Lipopan F did not constitute prior art since it "was not disclosed or available until October, 2001," after the claimed invention by Danisco of Lipase 3. (D. Mem. 36, quoting Novozymes Ex. 41, at 2.) Novozymes asserts that Danisco committed fraud in how it presented the Novozymes patents to the Patent Office.

This controversy involves a complex set of issues about what different parties knew and when they knew it. Certain facts now appear to be undisputed. Danisco concedes that the Novozymes patent documents disclose an enzyme that is in fact Lipopan F (although it was not so denominated in the patent documents) and that the use of Lipopan F in dough "satisfies all of the requirements of the asserted claims of the [Lipase 3] patent." (P. Mem.31.) It is also undisputed, however, that Danisco did disclose the existence of these patent documents to the Patent Office, though Novozymes claims that the disclosure was belated. (D.Mem.36.) In fact, Danisco actually withdrew its Lipase 3 application, after the patent examiner had already issued a Notice of Allowance,^[4] so as to provide the Novozymes patents to the examiner for consideration. (P. Mem.31-32.) It is difficult to see this as evidence of bad faith, let alone fraud.^[5]

However, Danisco took the position that the Novozymes patents did not "teach or suggest the instant invention." (D. Mem. 2, quoting Novozymes Ex. 34, at 1.) In effect, translated from the patent-ese, Danisco argued to the Patent Office that the Novozymes patent documents themselves did not disclose that the enzyme in question worked against lipids other than phospholipids, and did not refer to the use of the enzyme in dough. Therefore, even if it is now known that the enzyme referred to in the Novozymes patents is the product marketed as Lipopan F, and that Lipopan F has the triple lipase action in dough that was the supposed novel quality of Danisco's Lipase 3, no one would have known this from the Novozymes patent documents at the time of Danisco's claimed invention. (P. Mem.35-37.)

Novozymes attacks this theory, arguing that it shows a "fundamental misconstruction of the well-established law of inherency." (Reply 15.) Perhaps it does. But whether or not Novozymes is correct that the Novozymes patent documents, "together with the inherent properties of what they disclose, can (and do) make the [Danisco Lipase 3] claims unpatentable/invalid" (id.), Danisco's primary defense of the validity of its patent lies elsewhere.^[6]

The materiality of the parties' respective claims about what the Novozymes patent documents would have meant to one skilled in the art at the time the references came to light appears to be limited by the fact that upon disclosing the patents to the Patent Office, Danisco submitted a declaration from its inventors purporting to establish that Danisco's Lipase 3 had been invented before the effective date of the Novozymes patents. (P. Mem.31-35.) Since the Novozymes patents were entitled to priority only as of March 4, 1997 (about a month before Danisco's application was made), they would not be entitled to be considered "prior art" if Danisco's invention was made earlier than that date, and the Danisco inventors claims that it was. (Id.) Interestingly, for all its claims of fraud, Novozymes does not accuse the Danisco inventors of falsifying the affidavits and other evidence they submitted purportedly demonstrating that their discovery antedated the Novozymes patents.

Once again, Novozymes is correct to point out that even if Danisco's efforts to "swear behind the prior art" would have secured the patent regardless of what properties were or weren't inherent in the disclosures made in the Novozymes patent documents (a point Novozymes disputes, on grounds that lead even deeper into the arcana of Patent Office regulations), that would not entitle Danisco to fraudulently misrepresent the nature of those patents. (Reply 15.)^[7] But enough has been said to illustrate the complex issues that will have to be faced in this litigation in deciding the validity of the parties' competing claims. However those issues will ultimately be decided, on a full record, by the appropriate fact-finder, it cannot be said now, on the basis of the evidence presented by Novozymes on this motion, not merely that Novozymes is correct about some of these matters, but that it is so patently correct that Danisco's arguments to the contrary, when made to the Patent Office, must have constituted knowing fraud.^[8]

In short, whoever ultimately proves to have the better of the complex issues disputed by the parties, Novozymes has not succeeded in establishing any reasonable basis for a belief that any of the positions taken by Danisco in prosecuting this patent were taken in bad faith, let alone that they evidence a fraud such that all of Danisco's communications with its attorneys should be laid bare on the ground that they are communications in furtherance of a fraud.^[9]

CONCLUSION

For the foregoing reasons, Novozymes's motion to overrule Danisco's assertion of privilege and compel production of documents withheld on that ground is denied.

SO ORDERED.

NOTES

[1] Neither side presents sufficient information for a non-specialist judge to form an opinion of whether this level of activity is meaningful or not.

[2] The word "not" appears in Novozymes's memorandum, as it does in the text here, emphasized by italics, boldface, and underlining, all supplied by Novozymes (although the memorandum does not so indicate). The Court gets the point.

[3] As an attachment to its reply memorandum, Novozymes submits pages purporting to be minutes of two meetings attended by Dr. Sø05DFe, at which it was indicated that Lipase 3 is inactive against phospholipids. (Novozymes Exs. 43, 44.) It is wholly unclear that these documents, which consist of what appears to be e-mailed and cut-and-pasted text, are a complete and accurate description of what transpired at those meetings; and it is impossible to determine from them what statements were made by whom and in what context. For these reasons, the Court gives no weight to these documents. In any event, evidence that Dr. Sø05DFe attended meetings at which some presenters questioned whether Lipase 3 worked against phospholipids shows at best that the matter was subject to dispute, not that Dr. Sø05DFe or Danisco as an entity believed when they submitted the patent application that the enzyme did not do what it was claimed to do.

[4] As Novozymes observes, "[a] `Notice of Allowance' notifies a patent applicant that the pending claims in his application are `allowed,' i.e., ready to granted in a patent, if the appropriate remaining formalities are timely addressed, including payment of all Governmental fees." (D. Mem. 18 n. 23.)

[5] In particular, it is difficult to understand why, if Danisco had known about these references for two years prior to their actual disclosure to the Patent Office, as Novozymes claims (D.Mem.18-19), Danisco would have suppressed this knowledge until after the patent examiner had issued a Notice of Allowance, only then to withdraw its application, disclose the references to the Patent Office, and undertake to rebut their relevance. This behavior is not consistent with a fraudulent effort to hide the Novozymes patents from the patent examiner.

[6] In addition, whether or not the "inherency" doctrine applies such that the information claimed to be withheld by Danisco was material and should have been disclosed, is simply a dispute about the proper application of a legal standard to the facts of this case, not grounds for an allegation that Danisco intended to or did defraud the Patent Office.

[7] On the other hand, the crime-fraud exception requires that the fraud be material to the decision ultimately made by the Patent Office, and that would not be the case if the Office rejected the Novozymes patent documents as reflecting prior art on the basis that the Danisco invention occurred first. (D.Mem.26-29.) See, e.g., In re Spalding Sports World wide, Inc., 203 F.3d 800, 807-08 (Fed.Cir. 2000) (noting claim of fraud requires proof of a "justifiable reliance upon the misrepresentation by the party deceived which induces him to act thereon").

[8] Novozymes also points out that in challenging Novozymes's European patent application for Lipopan F, Danisco suggested that Lipopan F is identical to a lipase in a patent issued in 1985 (before Danisco's Lipase 3 invention), and that a skilled person would have known at the time that that identical lipase could have applications in the baking industry. Danisco argues primarily that it had no duty to disclose the position it took in the European proceedings, because its arguments (and the underlying facts) were immaterial to the issues before the Patent Office in passing on the '346 Patent. This dispute, like the rest, reveals parties taking adversary positions about legal standards and hotly-disputed facts relevant to the issues of materiality and the duty to disclose, not fraud or bad faith. (D. Mem. 22-24; P. Mem. 38-42; Reply 18-20.)

[9] For essentially the same reasons, the Court rejects Novozymes's request that the Court undertake an in camera review of the documents withheld by Danisco. (D.Mem.39-40.) Novozymes has failed to establish a "factual basis adequate to support a good faith belief by a reasonable person" that in camera review of the documents at issue "may reveal evidence to establish to claim that the crime-fraud exception applies." United States v. Zolin, 491 U.S. 554, 572, 109 S. Ct. 2619, 105 L. Ed. 2d 469 (1989) (quotation marks and citation omitted).