Illumina Inc. v. BGI Genomics Co., Ltd.

Opinion

ILLUMINA INC., et al., Case No. 20-cv-01465-WHO Plaintiffs, ORDER GRANTING MOTION TO v. AMEND BGI GENOMICS CO., LTD., et al., Re: Dkt. Nos. 233, 241, 244 Defendants. Defendants BGI Genomics Co., LTD., BGI Americas Corp., MGI Tech Co., Ltd., MGI Americas, Inc., and Complete Genomics, Inc. (“BGI”) move for leave to amend to add a new inequitable conduct defense. Plaintiffs Illumina, Inc. and Illumina Cambridge Ltd., (“Illumina”) oppose the motion, arguing that BGI’s proposed defense is futile, that BGI failed to act diligently in seeking leave, and that amendment would be prejudicial to Illumina. In light of the liberal amendment standard, and for the reasons set forth below, BGI’s motion is GRANTED. In addition, Illumina’s request to file a sur-reply is GRANTED. BGI’s motion to seal documents associated with its reply brief is DENIED. When prosecuting the ’444 patent, Illumina originally listed claims directed to a nucleotide or nucleoside modified to include an azidomethyl group bound to and blocking the 3’OH of the nucleotide or nucleoside. Dkt. No. 233 (“Mot.”) at 1. The patent examiner rejected the claims based on a paper published by Zavgorodny in 1991 (“Zavgorodny 1991”), which disclosed a method for placing an azidomethyl group on the 3’OH of a nucleoside. Mot. at 1; Dkt. No. 233-5 them to modified nucleotides and asserted that Zavgorodny 1991 does not teach or anticipate the claimed nucleotides. Dkt. No. 233-6 (Milowic Decl., Ex. 5”). Following this amendment, the patent examiner granted the ’444 patent, including claims 1 and 3, directed to a modified nucleotide. In its proposed Corrected First Amended Answer (“CFAA”), BGI seeks to allege a new inequitable conduct affirmative defense based on its claim that during this patent prosecution process, Illumina intentionally withheld or failed to disclose a reference to a paper by Terez Kovacs and Laslo Otvos title Simple Synthesis of 5-Vinyl and 5-Ethynyl- 2’ Deoxyuridine- 5’- Triphosphates, and published in Tetrahedron Letters, Vol. 29, pp 4525-4528, 1988 (“Kovacs”). See Dkt. No. 233-7 (“Kovacs”). BGI alleges that Kovacs discloses a methodology for converting nucleosides to nucleotides that is very similar to the method that two of the inventors of the ’444 patent, Drs. Xiaohai Liu and XiaoLin Wu, were using to create modified nucleotides. Dkt. No. 233-1 (“CFAA”) ¶ 337. It further alleges that a former Solexa employee named Sarah Lee, who worked in the same lab as Drs. Liu and Wu and who worked with them on converting nucleosides to nucleotides, made a hand notation that appears to reference the Kovacs publication in an April 2001 notebook entry. Id. ¶¶ 332-335. Although the handwritten note is hard to decipher, it appears to state “Tet. Let. 1998, 29, 4525.” Dkt. No. 233 (“Mot.”) at 2-3. The notation matches the publication information for the Kovacs paper, except that the date is 10 years off. See Kovacs at 1. BGI alleges that the Kovacs reference is material, both because it discloses the same methodology Liu and Wu used to convert nucleosides to nucleotides and also because it provides a motivation for a person of ordinary skill in the art (“POSITA”) to convert nucleosides to nucleotides. CCFA ¶ 338. Specifically, Kovacs notes that certain “nucleoside analogues” have proven useful as antivirals against herpes simplex virus infections and that “[i]nvestigations of the mechanisms by which these nucleoside analogues interfere with the cellular metabolism [ ] require the chemical synthesis of phosphorylated derivatives.” Kovacs at 4525. BGI asserts that “Kovacs teaches that converting nucleosides to nucleotides (a nucleoside with phosphates added) is 340. It alleges that based on Kovacs, a POSITA would have been motivated to use the method Kovacs discloses for converting nucleosides to nucleotides, to convert the 3’OH blocked azidomethyl nucleoside in Zavgorodny into a nucleotide. Id. ¶ 341. The Solexa notebook in which the alleged Kovacs reference appears was produced to BGI in April 2020. Dkt. No. 239 (“Opp.”) at 4. It is one of over 300 notebooks that have been produced to BGI. Dkt. No. 233-1 (“Milowic Decl.”) ¶ 2. BGI discovered the notation, and then the Kovacs paper, while preparing for the December 10, 2020 deposition of Dr. Wu. Id. ¶ 3. BGI filed its motion to amend on December 9, 2020, the last day to do so under the parties’ proposal that any amendment should happen no later than fifteen days after the Markman ruling in this case. See Dkt. No. 194 at 17. Federal Rule of Civil Procedure 15(a) allows a party to amend its pleading once within: (1) 21 days after serving the pleading or (2) 21 days after the earlier of service of a responsive pleading or service of a Rule 12(b) motion. Fed. R. Civ. P. 15(a). Outside of this timeframe, “a party may amend its pleading only with the opposing party’s written consent or the court’s leave.” Id. A court “should freely give leave when justice so requires.” Id. “Although the rule should be interpreted with ‘extreme liberality,’ leave to amend is not to be granted automatically.” Jackson v. Bank of Hawaii, 902 F.2d 1385, 1387 (9th Cir. 1990) (citation omitted). A court considers five factors in determining whether to grant leave to amend: “(1) bad faith, (2) undue delay, (3) prejudice to the opposing party, (4) futility of amendment; and (5) whether plaintiff has previously amended his complaint.” In re Western States Wholesale Nat. Gas Antitrust Litig., 715 F.3d 716, 738 (9th Cir. 2013) (quoting Allen v. City of Beverly Hills, 911 F.2d 367, 373 (9th Cir. 1990). “Prejudice to the opposing party is the most important factor.” Jackson, 902 F.2d at 1387. To state a claim for inequitable conduct, a party must allege that “(1) an individual misrepresentation of a material fact, failed to disclose material information, or submitted false material information; and (2) the individual did so with a specific intent to deceive the PTO.” Exergen Corp. v. Wal-Mart Stores, Inc., 575 F.3d 1312, 1327 n.3 (Fed. Cir. 2009) (internal citation omitted). Allegations of inequitable conduct must be pleaded with particularity pursuant to Federal Rule of Civil Procedure 9(b), which requires that the pleadings “identify the specific who, what, when, where, and how of the material misrepresentation or omission committed before the PTO.” Id. at 1328. To meet the intent prong, the pleading “must include sufficient allegations of underlying facts from which a court may reasonably infer that a specific individual (1) knew of the withheld material information or of the falsity of the material misrepresentation, and (2) withheld or misrepresented this information with a specific intent to deceive the PTO.” Id. at 1328–29. I have held that at the pleading stage, an inference of deceptive intent must be reasonable and drawn from the allegations of underlying fact. See Finjan, Inc. v. Check Point Software Techs., Inc., No. 18-cv-02621-WHO, 2019 WL 330912, at *4 (N.D. Cal. Jan. 25, 2019). “A reasonable inference is one that is plausible and that flows logically from the facts alleged, including any objective indications of candor and good faith.” Exergen, 575 F.3d at 1329 n.5. BGI moves for leave to amend to add a new inequitable conduct theory. See Mot. Illumina opposes the motion on three bases: (1) BGI did not act diligently; (2) amendment would be prejudicial to Illumina; and (3) amendment would be futile, focusing the bulk of its argument on futility. Opp. at 6-13. For the reasons discussed below, I conclude that BGI has acted with reasonable diligence, amendment would not be unduly prejudicial to Illumina, and BGI’s proposed inequitable conduct claim is not clearly futile under the relevant pleading standards. Accordingly, BGI’s motion for leave to amend is GRANTED.1 1 Illumina seeks leave to file a sur-reply, arguing that BGI’s reply brief raises a number of issues not included in its original motion for leave, including incorrect assertions of fact regarding the prior art references Illumina submitted during the ’444 patent application process. See Dkt. No. 244. BGI opposes the motion, arguing that Illumina has failed to follow the Northern District Local Rules and that a sur-reply is not appropriate. See Dkt. No. 258. BGI does not dispute, however, that its reply includes new and incorrect assertions of fact regarding whether Illumina I. DILIGENCE Illumina opposes BGI’s motion for leave because BGI did not diligently seek leave to amend. Opp. at 12-13. I disagree; BGI acted with reasonable diligence. BGI and Illumina agree that the relevant notebook in which BGI discovered the reference to Kovacs was produced in April 2020 and that more than 300 notebooks have been produced in this case. See Milowic Decl. ¶ 2; Opp. at 4. BGI argues that, given the large number of notebooks produced with handwritten notes – which are not computer searchable – and the relative illegibility of the relevant notation, it is reasonable that BGI did not discover the Kovacs reference until it was preparing for Dr. Wu’s deposition set for early December. Mot. at 4. Illumina counters that BGI did not act diligently, emphasizing that BGI had the notebook for over seven months before seeking leave to amend at the very last minute. Opp. at 12. BGI’s motion is based primarily on the discovery of a single, hard-to-read, handwritten notation in a notebook belonging to a non-inventor of the ’444 and ’973 patents. It is reasonable that it would take some time for BGI to manually review more than 300 notebooks and that BGI might prioritize the notebooks belonging to the patent inventors as part of its review process. It is also reasonable that it would take BGI some time to decipher or discover the relevance of the particular notation at issue here, given that the note is difficult to make out, the reference – if it is to Kovacs – lists the wrong publication year, and BGI was not previously focused on the Kovacs publication as relevant prior art. For these reasons I conclude that BGI has acted with reasonable diligence and any delay in seeking leave to amend is not a sufficient basis to deny leave. II. PREJUDICE Illumina also argues that amendment would be unduly prejudicial to Illumina because if amendment is granted, it will need to conduct additional investigation into Sarah Lee and her role, if any, in the prosecution of the ’444 patent. Opp. at 13. It further notes that BGI may seek further discovery on this issue, including, possibly, a deposition of Lee. Id. While allowing BGI to add its proposed inequitable conduct defense may result in additional costs to Illumina, and possibly additional discovery, these potential costs do not appear so substantial as to be unduly burdensome. Accordingly, I do not believe that amendment would cause significant prejudice to Illumina. Illumina spends the bulk of its opposition and sur-reply arguing that BGI’s proposed inequitable conduct claim is futile. See Opp. at 6-12; Dkt. No. 244-1 (“Sur-Reply”) at 2-6. Specifically, it argues that BGI has failed to adequately plead intent or materiality, two key requirements to an inequitable conduct defense. Opp. at 6. While I agree with Illumina that there are weaknesses to BGI’s proposed defense, I do not agree that the claim is futile at the pleading stage. A. Rule 9(b) Illumina does not dispute that BGI’s allegations are sufficiently particular to meet the heightened pleading standard of Rule 9(b). BGI has adequately alleged “the specific who, what, when, where, and how of the material misrepresentation or omission committed before the PTO” as required under Rule 9(b). Exergen, 575 F.3d at 1328. It alleges that Drs. Liu and Wu, two inventors of the ’444 and ’973 patents who had a duty of candor as inventors, did not disclose Kovacs as relevant prior art to the patent office at any point during the prosecution of the ’444 and ’973 patents. CFAA ¶ 347. BGI also asserts specific facts to support its claim that Drs. Wu and Liu were aware of the Kovacs reference, specifically that (1) their lab mate Sarah Lee made a reference that appears to be to Kovacs in a April 2001 notebook entry; and (2) Liu, Wu, and Lee all used a methodology to convert nucleosides to nucleotides that is remarkably similar to the methodology disclosed in Kovacs. Id. ¶¶ 332-335. BGI further states that Kovacs is material for two reasons: (1) because it teaches a specific methodology for converting nucleosides to nucleotides; and (2) because it provides a motivation for converting nucleosides to nucleotides, noting that doing so makes it possible to investigate the mechanism by which certain nucleoside antivirals work. Id. ¶¶ 337-339. And it explains how these facts would have allegedly impacted the patent examination process, alleging motivations to convert nucleotides [sic] into nucleotides as described in Kovacs, as well as a specific synthesis method for doing so, the Examiner would not have allowed a claim whose only difference from the disclosure in the Zavgorodny reference was that Zavgorodny disclosed azidomethyl blocking the 3’ OH of a nucleoside whereas claims 1 and 3 of the ’444 patent had been narrowed to cover azidomethyl blocking the 3’ OH of a nucleotide (a nucleoside with phosphates added).” Id. ¶ 343. Finally, BGI specifies that the failure to disclose Kovacs impacted the patent examiner’s assessment of claims 1 and 3 of the ’444 patent and claims 1 and 13 of the ’973 patent, all of which claim modified nucleotides having a blocking group, Z, C(F’)2—N3, where Z is azidomethyl. Id. ¶ 344. These allegations are specific and disclose the who, what, when, where, and how of its alleged inequitable conduct theory. They are sufficient to meet the particularity requirements of Rule 9(b). B. Knowledge & Specific Intent Illumina argues that BGI’s inequitable conduct claim fails to adequately allege specific intent under the Federal Circuit’s Therasense decision, which requires that intent to deceive the patent office be “the single most reasonable inference able to be drawn from the evidence.” Opp. at 5-7. While I agree that BGI’s allegations likely would not meet the standard set by Therasense, I have previously held, as have several other courts, that the Therasense standard does not apply at the pleading stage. See iLife Techs. Inc. v. AliphCom, No. 14-CV-03345-WHO, 2015 WL 890347, at *4 n.1 (N.D. Cal. Feb. 19, 2015) (“I decline to impose a standard that requires more than that set forth in Exergen, and will not apply Therasense at this stage of the proceedings.”). Accordingly, the relevant question is whether BGI’s proposed allegations are sufficient under the standard set forth in Exergen. Under Exergen, a party’s pleading “must include sufficient allegations of underlying facts from which a court may reasonably infer that a specific individual (1) knew of the withheld material information or of the falsity of the material misrepresentations, and (2) withheld or misrepresented this information with a specific intent to deceive the PTO.” Exergen, 575 F.3d at 1328. BGI’s allegations are sufficient to support a plausible inference that Drs. Liu and Wu had knowledge of the Kovacs publication. As noted above, BGI has alleged that a reference to Kovacs appears in the notebook of Liu and Wu’s lab mate Sarah Lee; that Liu, Wu, and Lee were all using the same methodology to convert nucleosides to nucleotides; and that this methodology is virtually identical to the methodology disclosed in Kovacs. CFAA ¶¶ 335-339. Reading these facts in the light most favorable to BGI, as is appropriate at the pleading stage, these allegations are sufficient to support a plausible inference that Liu and Wu were aware of the Kovacs reference. Another plausible inference is that Liu and Wu simply learned the methodology from Lee, or another scientist, and did not know of its origin. But at this stage, the relevant question is whether the facts could support a plausible inference of knowledge. I conclude that they can. It is a closer question whether BGI’s allegations support a plausible inference of specific intent to deceive, but I ultimately conclude that, on a permissive pleading standard, they do. As Illumina correctly notes, “[i]ntent to deceive cannot be inferred solely from the fact that information was not disclosed.” Optium Corp. v. Emcore Corp., 603 F.3d 1313, 1321 (Fed. Cir. 2010). But as the Federal Circuit has recognized, “direct evidence of deceptive intent is rarely available, and that intent may be inferred from circumstantial evidence.” Id. “A reasonable inference [of specific intent] is one that is plausible and that flows logically from the facts alleged, including any objective indications of candor and good faith.” Exergen, 575 F.3d at 1329 n.5. While BGI’s allegations of specific intent are relatively sparse, it has not simply alleged that Liu and Wu were aware of Kovacs and failed to disclose it. Instead, it alleges that Liu and Wu copied the specific methodology used in Kovacs for converting nucleosides into nucleotides, but did not disclose Kovacs as relevant prior art. CFAA ¶ 337. It has also asserted that in creating the ’444 patent, Illumina used a method very similar to a method disclosed in Zavgorodny 1991 for placing an azidomethyl group on the 3’OH of a nucleoside. Id. ¶¶ 327-328. BGI argues that Illumina deceptively failed to make clear in its ’444 patent application that this method came from Zavgorodny 1991, and instead presented the method as original. Dkt. No. 241-4 at 8.2 In sum, 2 BGI argues at length in its reply that Illumina engaged in a pattern of concealment and failed to disclose Zavgorodny 1991 or 2000 to the patent examiner until after the ’444 patent application was initially denied based on a reference to Zavgorodny 1991. Dkt. No. 4-8. These allegations are not included in BGI’s CFAA. Further, in its sur-reply, Illumina disputes BGI’s recounting of the prosecution history, and attaches a copy of Illumina’s information disclosure statement from June 2005, which appears to show that Zavgorodny 2000, which cites to Zavgorodny 1991, was BGI has alleged that Illumina borrowed two specific methodologies that were key to creating the claimed nucleotides, did not make clear to the patent examiner that these methodologies had been adopted from Zavgorodny 1991 and Kovacs, and did not disclose Kovacs at all as prior art. Although these allegations are no smoking gun, similarly sparse allegations have been found sufficient, at the pleading stage, to permit an inference of specific intent. See Capella Photonics, Inc. v. Cisco Systems, Inc., 77 F. Supp. 3d 850, 864 (N.D. Cal. 2014) (specific intent adequately pleaded where defendant alleged that plaintiff “initially refused to identify the precise flaw in its original patents when seeking reissue,” “failed fairly to disclose” a key prior art reference in a filing likely to be reviewed by the patent examiner, and defendants had significant financial incentives to deceive patent office); Coolsystems, Inc. v. Nice Recovery Systems LLC, Case No. 16-cv-02958-PJH, 2016 WL 6091577, *4 (N.D. Cal. Oct. 19, 2016) (specific intent adequately pleaded where defendant alleged that relevant prior art – which was allegedly virtually identical to claimed invention - was “buried” on 8th page of a list of 176 other minimally relevant references). I conclude that BGI’s allegations are sufficient, at this stage, to support a plausible inference of specific intent. C. Materiality Illumina argues that BGI’s proposed inequitable conduct defense is also futile because the Kovacs reference is cumulative of other prior art disclosed to the patent examiner, and is thus not material. I have previously held that it is not appropriate at the pleading stage to determine the materiality of prior art allegedly withheld from the patent office. See iLife Techs., 2015 WL 890347, at *6 (N.D. Cal. Feb. 19, 2015) (concluding that “an analysis of the European references’ materiality as prior art is not appropriate at this stage”); see also Breville Pty Ltd. v. Storebound LLC, No. 12-CV-01783-JST, 2013 WL 1758742, at *6 (N.D. Cal. Apr. 24, 2013) (“The Court does not read Exergen as requiring a court to determine, on the face of the pleadings, whether the allegedly missing material is in fact material and not cumulative.”). Illumina argues that many alleged failure to disclose Zavgorodny at the outset of prosecuting the ’444 patent. However, courts do consider materiality at the pleading stage and that I considered materiality with regard to BGI’s prior inequitable conduct defense. See Sur-Reply at 4 (citing Illumina, Inc. v. BGI Genomics Co., No. 19-CV-03770-WHO, 2020 WL 571030, at *2-4 (N.D. Cal. Feb. 5, 2020)). However, in reviewing BGI’s prior defense, my assessment of materiality focused on whether the allegedly misleading statements were material to the PTAB and Federal Circuit’s decisions based on those courts’ written analyses. Id. at 3. I did not attempt to assess whether a specific prior art reference would be material or cumulative in the patent prosecution process, a much more technical and fact-specific inquiry. Id. While it may be appropriate to consider materiality in some contexts, I decline to engage in a fact-intensive inquiry into the materiality of Kovacs as prior art at this stage. BGI’s allegations are sufficient to support a plausible inference that the Kovacs reference is material and not cumulative. It asserts that Kovacs discloses a motivation for converting nucleosides to nucleotides because “Kovacs teaches that converting nucleosides to nucleotides (a nucleoside with phosphates added) is necessary to study the mechanism that normally occurs in a cell, in a test tube instead.” CCFA ¶ 340. Although BGI previously argued that Zavgorodny 1991 also disclosed this motivation, Illumina does not dispute that the references are different. BGI contends that the Kovacs discussion on this point is more detailed than any other prior art that was submitted to the patent office, including Zavgorodny, and is thus not cumulative. See Reply at 11- 12. As noted above, it is not appropriate at this stage to make a factual determination whether Kovacs is or is not cumulative of the other prior art references that were before the patent office. BGI has made facially plausible allegations that Kovacs discloses something more and different than these other references. At this stage, these allegations are sufficient to allege materiality. For these reasons, I conclude that BGI’s proposed inequitable conduct defense is not clearly futile on a permissive pleading standard. Accordingly, BGI’s motion for leave to amend is BGI has filed a motion to seal portions of its reply brief and supporting exhibits. See Dkt. 1 because they have been designated confidential by Illumina and takes no position on whether the 2 documents should be sealed. Dkt. No. 241-2. Pursuant to Local Rule 79-5(e), as the designating 3 party, Illumina had four days from the date BGI filed its motion to submit a substantive 4 declaration providing justification for sealing these documents. [lumina did not file such a 5 declaration. In addition, upon review, these documents do not appear to contain sensitive 6 information suitable for sealing. Accordingly, BGI’s motion to seal is DENIED. The sealed 7 documents will be unsealed on the docket. 9 For the reasons discussed above BGI’s motion to amend is GRANTED. BGT is instructed 10 to file its proposed CFAA, currently filed at Docket No. 233-2, as a separate document on the 11 docket. In addition, I!lumina’s motion to file a sur-reply is GRANTED. BGI’s motion to seal 12 documents associated with its reply brief is DENIED and the provisionally sealed documents will 13 be unsealed on the docket. IT IS SO ORDERED. 15 Dated: February 8, 2021

S 18 William H. Orrick United States District Judge 19 20 21 22 23 24 25 26 27 28