NeuroGrafix v. Brainlab, Inc.

Opinion

IN THE UNITED STATES DISTRICT COURT FOR THE NORTHERN DISTRICT OF ILLINOIS EASTERN DIVISION

NEUROGRAFIX, et al., ) ) Plaintiffs, ) ) vs. ) Case No. 12 C 6075 ) BRAINLAB, INC., et al., ) ) Defendants. )

ORDER ON DEFENDANTS' MOTION TO EXCLUDE (dkt. no. 479) AND MOTION TO QUASH TRIAL SUBPOENAS (dkt. no. 492)

A jury trial in this patent infringement case began on November 3, 2020, but on plaintiffs' motion the Court declared a mistrial a day later, after one of the eight jurors apparently became ill. For the retrial, which is set for April 26, 2021 (but is about to be moved to May 3, 2021 as the Court has previously discussed), defendants have moved to preclude plaintiffs from pursuing what they have termed their "numerosity" damages theory and to preclude plaintiffs from arguing or introducing evidence before the jury regarding claimed discovery violations by defendants. Defendants have also moved to quash subpoenas served by plaintiffs regarding two witnesses associated with defendants. The Court rules on these motions in this Order. 1. Motion to exclude "numerosity" damages / claimed discovery violations a. Plaintiffs' damages computations What plaintiffs refer to as "numerosity" means the following (as plaintiffs describe it): "The word 'numerosity' simply means—how many instances of infringement occurred." Pls.' Opp. to Brainlab's Mot. to Exclude (dkt. no. 484) at 12. During discovery, plaintiffs' expert submitted a report disclosing a reasonable royalty damages figure of $310,000. But in their submission for the final pretrial order—which they submitted around a month before trial was set to begin—plaintiffs stated that they would seek $180 million in a usage-based reasonable royalty; $150 million consisting of a reasonable royalty on defendants' sales of "BrainSuites"; and lost profits in an

unspecified amount. Defendants' current motion is focused on the first component, as the others have been the subject of earlier rulings. In their opening statement at trial, plaintiffs increased the $180 million royalty figure still further: their counsel argued for a reasonable royalty figure well in excess of $300 million. Defendants' motion to exclude is based largely on disclosure issues. Federal Rule of Civil Procedure 26(a)(1) requires a party to disclose, among other things, "a computation of each category of damages claimed by the disclosing party." Fed. R. Civ. P. 26(a)(1)(A)(iii). Plaintiffs' Rule 26(a)(1) disclosures did not specify an amount of damages, a computation, or a method of computation; plaintiffs stated that they would provide a computation later, after obtaining discovery from defendants. In their later

responses to defendants' interrogatories served in February 2017, plaintiffs likewise provided no computation. Defs.' Mem., Ex. B at 6-7. A later supplement to plaintiffs' interrogatory answers evidently made reference to plaintiffs' second amended complaint filed in June 2017. This document likewise did not include any reasonable royalty figure but said that a royalty "should be based on the value of the system sold." Defs.' Mem., Ex. C ¶ 30. The first actual royalty figure or computation provided by plaintiffs was in a damages expert report by John Elmore served in September 2017; he calculated a reasonable royalty of $55,000 based on the sales price of defendants' software. More recently, in March 2020, Elmore served an amended report that claimed a usage-based reasonable royalty of $310,000. This was based on identifying a representative hospital; determining usage of DTI at the hospital; estimating the percentage of that usage that represented Brainlab's software based on information from a survey expert;

and then extrapolating to calculate an overall royalty amount. After March 2020, plaintiffs did not further supplement their Rule 26(a)(1) disclosures or their interrogatory answers, nor did they serve a further amended or supplemental expert report. Their first reference to a vastly increased figure was in their submission for the final pretrial order, first submitted in or about late September/early October 2020, in which plaintiffs provided a royalty figure of $180 million. In that submission, plaintiffs appeared to be adopting a different model from the one posited by Elmore. Specifically, they appeared to be basing their calculation, at least in part, on assumptions or inferences resulting from their contention that defendants had either failed to produce or destroyed evidence relevant to the determination of damages. In

this submission, plaintiffs referred to an anticipated motion for sanctions—which they had not yet filed—and asked the Court to defer entry of the final pretrial order until that motion had been filed and ruled upon. Plaintiffs did not file their motion for sanctions until October 29, 2021, the Thursday before the jury trial's Tuesday, November 3 starting date. They argued that defendants had withheld discovery regarding, among other things, customer usage and sales. Plaintiffs sought entry of a default judgment, or alternatively a monetary sanction, an order for production of the supposedly withheld records, a continuance of the trial, and leave to submit a new damages expert report. The Court ordered defendants to respond to the motion the next day (Friday, October 30), and it ruled on the motion on Sunday, November 1. The Court held that the motion was untimely and also rejected of plaintiffs' contentions on the merits. Specifically, the Court "conclude[d] that NeuroGrafix has not shown that Brainlab has withheld material that it was required to

produce and therefore [found] that NeuroGrafix's motion lacks merit." Order on Plaintiffs' Motion for Sanctions (dkt. no. 446) at 3. During opening statement at the jury trial on November 3, plaintiffs' counsel argued for an even higher royalty-based damage figure, $317 million—more than $130 million greater than the belated $180 million figure plaintiffs had included in their submission for the final pretrial order. In addition, despite the Court's ruling denying their motion for sanctions and concluding that their contentions in that motion lacked merit, plaintiffs' counsel argued in opening statement that plaintiffs had to estimate their damages because defendants had not produced records they were required to produce. In their motion, defendants ask the Court to preclude plaintiffs' higher damages

calculation(s) and theory on the ground of nondisclosure, and they also ask the Court to bar argument, evidence, or discussion of claimed discovery withholding or spoliation at trial. In opposition, plaintiffs argue that their damages contentions and data were disclosed. Plaintiffs cite, first, what they refer to as an interrogatory answer from 2017 that included the following statement: If BrainLAB had respected the patent, it would not have been able to provide ANY tractographic systems without making an arrangement with IBSC [Image-Based Surgicenter Corp., one of the plaintiffs]. In fact, IBSC had extensive discussions with BrainLAB competitors who ultimately judged that if BrainLAB was not licensing then they should not license. Therefore, the losses at any facility with a [sic] 1,000 scans per year was $4,000 x 1,000 or $4 million/year. Between 2009 and 2013, the loss to IBSC was $20 million at each such facility.

Pls.' Resp., Ex. A3. This reference is a bit confusing. Plaintiffs cite it as "NeuroGrafix response to Brainlab Interrogatory on Damages, provided by NeuroGrafix on August 14, 2017, and filed by Brainlab on Sept. 5, 2017 (see MDL Dkt. # 421-5, Sept. 5, 2017, pages 50/54, see Exh A3)." Pls.' Resp. at 3. MDL docket entry 421-5, however, is not on its face a response to a damages interrogatory. Docket entry 421, to which entry 421-5 is attached, is a motion for protective order filed by defendants in the MDL proceeding, regarding subpoenas that plaintiff had served on third parties. Docket entry 421-5 is exhibit D to an affidavit submitted by defendants' counsel in support of the motion for protective order. Defense counsel's affidavit says that docket entry 421-5— the item plaintiffs refer to as a damages interrogatory response—is a "Proposed Revised Amended Complaint that Neurografix has stated it intends to file in [sic] near future"—not a damages interrogatory response. Case No. 1:13-md-02432-RGS (D. Mass.), dkt. no. 421-1 ¶ 6. And that is exactly what docket entry 421-5 appears to be on its face: it is entitled "Proposed Revised Amended Complaint for Patent Infringement and Demand for Jury Trial." See id., dkt. no. 421-5 at 1. It does, however, appear to be verified under oath by plaintiffs' principal Dr. Aaron Filler. See id. at 54. And both sides seem to say that the proposed amended complaint somehow got incorporated into an

interrogatory answer by plaintiffs, see Defs.' Reply at 4, so the Court will treat it as such. The problem with the above-quoted statement from the proposed amended complaint/interrogatory answer is that plaintiffs quote (and attach) only part of the relevant paragraph and thus offer it out of context. See Pls.' Resp., Ex. A3 (providing, and highlighting, the last two sentences of paragraph 114). Here is the full paragraph 114 from MDL docket entry 421-5: As shown in the attached material Image-Based Surgicenter Corporation offered several different commercial arrangements (see Confidential Sealed Exhibit H). These included a design/build arrangement such as the system offered by BrainLAB. However, historically, working with its Neurography Institute partner, IBSC was experienced with providing complete scan services through the US and in Europe. For NIMA, each scan was paid at $4,000 per scan. It is readily seen from Exhibits I and J that thousands of scans are performed annually for use in intraoperative MRI by each of the larger FiberTracking facilities (e.g. Yale, Henry Ford Hospital). Additionally IBSC offered a supervisory radiological presence at $500 per scan. For these services, IBSC would monitor and supervise the technologist performing the scan— electronically in real time from its central offices with a pool of experts— then move the scan electronically to its workstations to perform tractographic analysis on an ongoing basis as the surgery progressed. If BrainLAB had respected the patent, it would not have been able to provide ANY tractographic systems without making an arrangement with IBSC. In fact, IBSC had extensive discussions with BrainLAB competitors who ultimately judged that if BrainLAB was not licensing then they should not license. Therefore, the losses at any facility with a 1,000 scans per year was $4,000 x 1,000 or $4 million per year. Between 2009 and 2013, the loss to IBSC was $20 million at each such facility.

Case No. 1:13-md-02432-RGS (D. Mass.), dkt. no. 421-5 ¶ 114. As defendants note, see Defs.' Reply at 5, this does not read as a description of a usage-based reasonable royalty loss, which is what plaintiffs are seeking now. Rather, it appears to read as a description of lost revenues or profits. The $4,000 figure in the second-last sentence refers back not to a royalty amount, but to what plaintiffs say NIMA, one of the plaintiffs, was paid for each scan performed. The $4,000 figure can't possibly be a per-use royalty figure; it's twenty-five times the $160-per-scan royalty figure that plaintiffs used in their opening statement at trial.1

1 It also bears repeating that when plaintiffs provided an expert report some time after this, the report adopted figures that bear no relationship to the quoted paragraph from the proposed amended complaint. In short, the key document that plaintiffs reference as long ago disclosing their current "numerosity" approach does not actually do that at all. And even if the Court is understanding paragraph 114 incorrectly, it does not suffice as a proper disclosure of a computation of the reasonable royalty damages that plaintiffs now appear to be

claiming. The illustration in paragraph 114 is provided only as an example, only involving one facility. And the amount it discloses, $20 million, bears no relationship at all to the figure(s) that plaintiffs claimed in the final pretrial order submission or in opening statement at trial. Under the circumstances, the document cited by plaintiffs cannot be understood as satisfying their obligation to disclose, during the period for discovery, the damages computation they are now seeking to advance. Plaintiffs also cite other instances where they disclosed, at least inferentially, large numbers of allegedly infringing uses at various hospitals. See Pls.' Resp. at 2-4, 6. In addition, plaintiffs cite the Rule 30(b)(6) deposition testimony of Joseph Doyle, a managing director of Brainlab, which they interpret as indicating that there close to 600

U.S. locations at which defendants' software was used. See id. at 5-6. (Doyle's actual testimony, which plaintiffs quote, referenced "around sixty/sixty-five a year" software license sales, and plaintiffs say that if this held for 2005 through 2013, there would be a total of 585. See id. at 5.) They also cite a published study from which they attempt to draw an inference regarding the number of craniotomies for brain tumors that are conducted in the U.S. See id. at 8-9. Plaintiffs further cite a declaration by Dr. Filler, see id. at 8 (citing dkt. no. 151 at 3), but that was not a royalty damages disclosure as such—it was an affidavit submitted on an issue relating to recovery of lost profits, on which the Court ruled against plaintiffs. Finally, the Court notes that in supporting their "numerosity" damages, plaintiffs expressly rely on defendants' purported "refus[al] to produce the data" regarding the number of systems it has in place and "fail[ure] to produce" data from a database called SalesForce. See id. at 6 (heading IV title), 7-8; id. at 10 (heading VII title), 10-11.

These are matters on which the Court has ruled against plaintiffs, in the aforementioned denial of their motion for sanctions (and the denial of their motion for reconsideration of that ruling). Toward the end of their response, plaintiffs explain their current royalty damages figure as "includ[ing] the following data points": • 657 FiberTracking locations–which they say "correlate[s]" with Doyle's testimony; • 1,300 "bundled" locations—which they say "correlate[s]" with a comment by Doyle; • 250 procedures per year—which they say is well below the figure they obtained

for one large hospital; • 7 years of inducement; • $160 per scan "based on reduced value of $3,200 x .05 per scan for the royalty base"; and • 0.58 "market share modified to remove the Siemens sites from the market." The result, plaintiffs say, is $317 million.2 See id. at 11-12. Plaintiffs concede that there is approximation in their figures but argue that reasonable royalty damages "necessarily

2 The Court notes that this computation is nothing at all like the formula that plaintiffs offered in the 2017 proposed amended complaint/interrogatory answer discussed earlier. This is one more reason why the 2017 document cannot appropriately be considered as a timely fulfillment of plaintiffs' damages disclosure obligations. involve[ ] an element of approximation and uncertainty." See id. at 14 (citing Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1336 (Fed. Cir. 2009)). Plaintiffs contend that the proposition that they were relying on "numerosity" and on software bundling was adequately disclosed in various ways. See id. at 12-13. They criticize Brainlab for

"wait[ing] until after the trial date to start seeking to exclude this method of analysis . . . ." Id. at 15. It may be true, as plaintiffs contend, that some or all of the pieces of their newly- disclosed damages calculation can be found or derived from information contained in the extensive record in this case. What is completely missing, however, is any attempt prior to the final pretrial order to put those pieces together into anything approaching a royalty damages computation. Plaintiffs do not argue otherwise. Specifically, they do not contend that they disclosed the $317 million figure (or anything like it) at any time before opening statement, nor do they contend that they disclosed the $180 million figure (or anything anywhere close to it) at any time before their submission of a draft of

the final pretrial order, long after the deadline for completing discovery. Plaintiffs did disclose a royalty damages calculation—the original and revised reports of their expert Elmore—but they basically left that in the dust, in favor of a figure literally 1,000 times greater. This is despite the fact that all of the information that plaintiffs cite (aside from the alleged discovery abuses by defendants) was available to them before Elmore's March 2020 revised report, in which he proposed a royalty calculation of $310,000. Were the Court to permit plaintiffs to alter their damages calculation in this way just before trial, long after the close of discovery, it would turn the Civil Rules' disclosure requirements on their head. As the Court has indicated, Rule 26(a)(1) required plaintiffs to disclose a "computation" of their damages. The Court does not fault plaintiffs for not doing that at or near the outset of the case when they first made their Rule 26(a)(1) disclosures, as they needed more information at that point. But when plaintiffs did disclose an actual "computation" of their royalty damages, the only way they did that

was via their expert's reports; they did not formally supplement their Rule 26(a)(1) disclosures. The expert's March 2020 report disclosed a computation that came out to $310,000. That, under the circumstances, was plaintiffs' highest "computation" of reasonable royalty damages under Rule 26(a)(1). The expert report was never amended to propose a larger figure. If one construes plaintiffs' $180 million figure in their submission for the pretrial order as an attempt to supplement their earlier disclosures, it was not "timely" as required by Rule 26(e)(1)(A). At that point defendants had no way to counter plaintiffs' newly-minted figure, as discovery (including expert discovery) had long since closed. As defendants point out, they had taken the depositions of plaintiffs' experts and had

obtained responsive expert reports based on what plaintiffs' experts had said. But then, at virtually the last minute, plaintiffs pulled the rug out from under the disclosure they had made via their experts' reports. In sum, the plaintiffs' $180 million and $317 million royalty figures, and the computations underlying them, were not timely disclosed as required under Rule 26(a)(1). Plaintiffs cannot take refuge in Rule 26(e)'s exception to the duty to supplement that applies when information has "otherwise been made known" to the opposing party "during the discovery process or in writing." Fed. R. Civ. P. 26(e)(1)(A). Plaintiffs may have identified where the bits and pieces that they are relying upon came from, but what's significant is how those bits and pieces are put together—the "computation" expressly required by Rule 26(a), which plaintiffs made no attempt to provide for their nine-figure royalty claim until far too late. It is not sufficient that, as plaintiffs seem to contend, defendants might have been able to figure it out from traces left at various

points in the extensive record in this case. See generally Morris v. BNSF Ry. Co., 969 F.3d 753, 766 (7th Cir. 2020) ("Complying [with required disclosures] should be a priority—not something brushed off as tedious or unimportant so long as the information disclosed late can somehow be unearthed like a needle in a haystack within a prior discovery production."). The Court also notes, again, that all of the information on which plaintiffs now rely was available to them at or around the time of Elmore's amended report submitted in March 2020. Yet it was not incorporated into his computation or any other disclosure provided to defendants. Under Federal Rule of Civil Procedure 37(c)(1), when a party fails to provide information as required by Rule 26(a) or (e), "the party is not allowed to use that

information" at trial, "unless the failure was substantially justified or is harmless." Fed. R. Civ. P. 37(c)(1). The sanction of exclusion "is automatic and mandatory unless the sanctioned party can show that its violation of Rule 26(a) was either justified or harmless." David v. Caterpillar, Inc., 324 F.3d 851, 857 (7th Cir. 2003). Plaintiffs' late disclosure of the damages computation upon which they now rely—both the final pretrial order version and the trial version—is neither justified nor harmless. To the extent plaintiffs contend it is justified by defendants' claimed failure to produce records, the Court has ruled against plaintiffs on that claim, so it cannot serve as a justification. And plaintiffs have offered no other viable justification. Moreover, plaintiffs' belated disclosure is not harmless because, as the Court has discussed, defendants have not had a fair opportunity to obtain and marshal information, including expert reports, to attempt to counter plaintiffs' computation or its component parts. The last-minute disclosure of a damages bottom line that was, initially, 600 times what plaintiffs' expert

had disclosed and then increased to 1,000 times the expert's figure quite plainly came as a surprise, and defendants had no ability to cure it, given the closure of fact and expert discovery. The Court acknowledges the importance of the damages calculation in this case, but that simply reinforces the proposition that was incumbent upon plaintiffs to disclose their computation in a timely fashion as required by the Rules and not to come in at the eleventh hour, point to a number of strands of inference, and say the conclusion was somehow there for the finding. It is no answer to this to say, as plaintiffs have suggested at various junctures, that any problem can be cured via a continuance (combined, perhaps, with granting plaintiffs leave to submit a new expert report). This would unreasonably disrupt the trial,

which has been set several times and started once, and which is now scheduled to start in a matter of weeks. Moreover, saying that a nondisclosure or unreasonably belated disclosure may be waved away via a continuance is tantamount to saying that the required disclosures and discovery deadlines just don't matter, because deadlines and dates, including trial dates, can always be extended or moved. The law does not require this—particularly in a case as old at this one. For these reasons, the Court grants defendants' motion to exclude and bars plaintiffs from using at trial the "numerosity" damages figures or computation(s) described in their final pretrial order submission or in their opening statement (and supporting slides) given in the November 2020 partial trial. b. Evidence/questioning/argument based on withholding of discovery The Court also bars plaintiffs from offering evidence, questioning, or argument to the effect that defendants withheld material that they should have produced during

discovery. As indicated, the Court ruled against plaintiffs regarding this contention on both their motion for sanctions and motion for reconsideration. Having lost on this point, plaintiffs cannot appropriately put these contentions before the jury. Jurors decide the facts from the evidence and apply the governing law provided by the Court to those facts; they do not decide discovery disputes. The Court understands that plaintiffs disagree with the Court's ruling on the sanctions motion, but at this point that is a matter for them to address on appeal, not before the jury at trial. 2. Motion to quash trial subpoenas Defendants have also moved to quash subpoenas served by plaintiffs for two Illinois-based witnesses associated with defendants: Andrea Harrison and Ann Marie

LaCasha. Defendants argue that the only basis upon which they could possibly be called is to elicit testimony regarding the allegedly withheld discovery. Defendants argue that because the Court has ruled against plaintiffs on that issue, there is no basis for them to call either Harrison or LaCasha. The Court agrees with defendants that the subjects of missing evidence, withheld discovery, and the like are inadmissible at trial, as the Court has just ruled. But defendants have not shown that there is no other proper purpose for which these witnesses may be called. For this reason, the Court declines to quash the subpoenas. Any objections to particular testimony that plaintiffs may seek to elicit from these witnesses must be asserted during their testimony. The Court reserves the right to revisit this ruling should it become apparent that plaintiffs are calling these witnesses to elicit testimony that is or has been ruled to be inadmissible. Conclusion For the reasons stated above, the Court grants defendants' motion to exclude [dkt. no. 479] and denies defendants’ motion to quash [dkt. no. 492]. Date: March 17, 2021

United States District Judge