ImmunoGen, Inc. v. Vidal

Full title: IMMUNOGEN, INC., Plaintiff, v. Katherine VIDAL, Under Secretary of…

Court: United States District Court, E.D. Virginia, Alexandria Division

Date published: Jan 30, 2023

653 F. Supp. 3d 258 (E.D. Va. 2023)

Opinion

Case No. 1:20-cv-274

2023-01-30

IMMUNOGEN, INC., Plaintiff, v. Katherine VIDAL, Under Secretary of Commerce for Intellectual Property and Director of the United States Patent and Trademark Office, Defendant.

Craig Crandall Reilly, Law Office of Craig C. Reilly, Alexandria, VA, Alexander Vincent Alfano, Pro Hac Vice, Ryan Elizabeth Conkin, Pro Hac Vice, John Christopher Rozendaal, Sterne, Kessler, Goldstein & Fox PLLC, Washington, DC, for Plaintiff.

T.S. Ellis, III, United States District Judge

Craig Crandall Reilly, Law Office of Craig C. Reilly, Alexandria, VA, Alexander Vincent Alfano, Pro Hac Vice, Ryan Elizabeth Conkin, Pro Hac Vice, John Christopher Rozendaal, Sterne, Kessler, Goldstein & Fox PLLC, Washington, DC, for Plaintiff.

FINDINGS OF FACT AND CONCLUSIONS OF LAW

T.S. Ellis, III, United States District Judge

In this civil action to obtain a patent under 35 U.S.C. § 145, Plaintiff ImmunoGen, Inc. ("ImmunoGen") seeks a judgment that ImmunoGen is entitled to a patent for the invention specified in the claims of U.S. Patent Application Number 14/509,809 (the "'809 Application"). The '809 Application describes a dosing regimen for administering a drug that treats ovarian and peritoneal cancer. Because ImmunoGen already received a patent for the drug specified in the '809 Application, which is known as IMGN853, the '809 Application seeks only to patent a dosing regimen for administering IMGN853. Specifically, the '809 Application seeks to patent a method of dosing IMGN853 based on patients' adjusted ideal body weight ("AIBW") instead of total body weight ("TBW").

After a patent examiner at the U.S. Patent and Trademark Office (the "PTO") rejected the '809 Application and the Patent Trial and Appeal Board (the "PTAB") affirmed the rejection, ImmunoGen brought this action against Defendant Katherine Vidal in her official capacity as the Under Secretary of Commerce for Intellectual Property and Director of PTO. Pursuant to 35 U.S.C. § 145, ImmunoGen now seeks a judgment setting aside the PTO's decision that the claims in the '809 Application are unpatentable and declaring that ImmunoGen is entitled to issuance of a patent for the invention claimed in the '809 Application. In response, the PTO argues that ImmunoGen cannot receive a patent for the '809 Application for three reasons: (1) the claims of the '809 Application are fatally indefinite; (2) the '809 Application is unpatentable because it is obvious in light of the prior art; and (3) the '809 Application is invalid under the doctrine of obviousness-type double patenting.

This matter came before the Court on November 1, 2022 for a three-day bench trial. There, the parties stipulated various undisputed facts and presented substantial

testimonial and documentary evidence. After the close of evidence on November 3, 2022, the parties submitted proposed findings of fact and conclusions of law. Thereafter, at a hearing on December 15, 2022, counsel for each party presented closing arguments and the matter was taken under advisement. The matter is therefore now ripe for disposition.

Set forth below, pursuant to Rule 52, Fed. R. Civ. P., are the Court's findings of fact and conclusions of law. In short, judgment must enter in favor of the PTO because the claims at issue in ImmunoGen's '809 Application are unpatentable as indefinite and obvious. Furthermore, the '809 Application is unpatentable under the doctrine of obviousness-type double patenting.

I. LEGAL PRINCIPLES

Before setting forth the Findings of Fact, it is appropriate to provide a brief description of the governing legal principles, which provide helpful context. These legal principles will be stated in further detail in the Conclusions of Law.

This civil action was brought under 35 U.S.C. § 145, which provides that "[a]n applicant dissatisfied with the decision of the [PTAB] ... may ... have remedy by civil action against the Director [of the PTO] in the United States District Court for the Eastern District of Virginia." Id. In a civil action pursuant to § 145, the district court has the "power to set aside any ruling refusing a patent and determine patentability de novo." Disney Enterprises, Inc. v. Rea, 940 F. Supp. 2d 288, 291 (E.D. Va. 2013) (quoting Hitachi Koki Co., Ltd. v. Doll, 620 F. Supp. 2d 4, 16 (D.D.C. 2009)). In a § 145 action, the dissatisfied patent applicant challenging the decision of the PTAB ordinarily bears the burden of proof by a preponderance of the evidence. See Dome Patent L.P. v. Lee, 799 F.3d 1372, 1377-79 (Fed. Cir. 2015). Where the PTO raises new arguments during the § 145 action that the PTO did not raise during patent prosecution or before the PTAB, however, the PTO bears the burden of proving the newly raised defense by a preponderance of the evidence. See Hyatt v. Iancu, 332 F. Supp. 3d 113, 120 (D.D.C. 2018) (reversed on other grounds) (explaining that the PTO may raise new affirmative defenses in a § 145 proceeding and that the PTO bears the burden of proof on such affirmative defenses). The patent examiner and the PTAB rejected the '809 Application on the grounds of obviousness and obviousness-type double patenting. After ImmunoGen initiated this § 145 litigation, the PTO raised the additional argument that the '809 Application is fatally indefinite.

With respect to indefiniteness, the Patent Act requires that a patent specification "conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor ... regards as the invention." 35 U.S.C. § 112(b) (emphasis added). Supreme Court precedent requires a "reasonable certainty" standard: a patent is invalid for indefiniteness "if its claims, read in light of the specification ... and the prosecution history, fail to inform, with reasonable certainty, those skilled in the art about the scope of the invention." Nautilus, Inc. v. Biosig Instruments, Inc., 572 U.S. 898, 901, 134 S.Ct. 2120, 189 L.Ed.2d 37 (2014). In determining whether a patent application is indefinite, the district court must first evaluate the language of the claims, then look to pertinent intrinsic evidence including the application's prosecution history, and finally evaluate extrinsic evidence of indefiniteness. Berkheimer v. HP Inc., 881 F.3d 1360, 1363-64 (Fed. Cir. 2018).

The Patent Act also requires that patents address "non-obvious subject

matter." 35 U.S.C. § 103. The Supreme Court has explained that in determining whether a patent application is obvious, district courts must make three factual inquiries. Graham v. John Deere Co. of Kan. City, 383 U.S. 1, 17, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966). The district court must first determine the level of ordinary skill in the art at the time the invention was made, then determine the scope and content of the prior art, and finally identify the differences between the prior art and the claims at issue. Id. If the "difference between the subject matter sought to be patented and the prior art" is "such that the subject matter as a whole would have been obvious at the time to a person skilled in the art," then the subject matter is obvious and cannot be patented. Id. at 15, 86 S.Ct. 684. The Supreme Court has also explained that a patent for an invention "composed of several elements is not proved obvious merely by demonstrating that each of its elements was, independently, known in the prior art." KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 418, 127 S.Ct. 1727, 167 L.Ed.2d 705 (2007). Instead, the factfinder must consider (1) whether a skilled artisan would have a motivation "to combine the teachings of the prior art references to achieve the claimed invention," and (2) whether the skilled artisan "would have had a reasonable expectation of success in doing so." Novartis Pharm. Corp. v. West-Ward Pharm. Int'l Ltd., 923 F.3d 1051, 1059 (Fed. Cir. 2019). As the final step in the obviousness analysis, district courts must look to objective "secondary considerations" including unexpected results, long-felt but unmet needs, failure of others, copying by competitors, and industry praise. Graham, 383 U.S. at 17-18, 86 S.Ct. 684; Leo Pharm. Prods., Ltd. v. Rea, 726 F.3d 1346, 1357-58 (Fed. Cir. 2013).

Finally, § 101 of the Patent Act forbids an individual from obtaining more than one patent on the same invention, which is known as "double patenting." Abbvie Inc. v. Mathilda & Terence Kennedy Inst. of Rheumatology Trust, 764 F.3d 1366, 1372 (Fed. Cir. 2014). For this reason, courts have created the doctrine known as obviousness-type double patenting, which prohibits a patent applicant from receiving a patent for claims which are obvious in light of earlier-issued patents.

II. FINDINGS OF FACT

The facts recited here are those found pursuant to Rule 52(a)(1), Fed. R. Civ. P., on the basis of the record as a whole, including a three-day bench trial. In the course of the bench trial, six witnesses testified in total—five for ImmunoGen and one for the PTO. ImmunoGen first presented three fact witnesses: Drs. Joseph Ponte and Robert Lutz, inventors of the '809 Application who testified as to the development of IMGN853 and the claimed dosing regimen, and Dr. David O'Malley, a medical doctor who testified about his involvement with the clinical trials of IMGN853. ImmunoGen also presented two expert witnesses: Dr. Anthony Tolcher, a clinical oncologist experienced in drug development, and Dr. Dhaval Shah, a professor of pharmaceutical sciences experienced in pharmacokinetics. Both of ImmunoGen's expert witnesses testified that the '809 Application is neither obvious nor indefinite. In response, the PTO presented one expert witness: Dr. William Figg, Associate Director of the Cancer Center at the National Cancer Institute and head of the Clinical Pharmacology Program at the National Institute of Health. Dr. Figg, who is experienced in the areas of clinical pharmacology and drug development, testified that the dosing regimen disclosed in the '809 Application is both obvious and indefinite. While ImmunoGen compensated all five of its witnesses for their time, the PTO's only witness, Dr. Figg, was not paid

by the PTO for the time he spent working on the case.

A. Parties

1. Plaintiff ImmunoGen, Inc. is a corporation organized and operating under the laws of the Commonwealth of Massachusetts with a place of business in Waltham, Massachusetts. ImmunoGen is the assignee of the '809 Application at issue in this case.

2. Defendant Katherine K. Vidal is sued in her official capacity as the Under Secretary of Commerce for Intellectual Property and Director of the United States Patent and Trademark Office.

B. Development of IMGN853 and the Dosing Regimen at Issue

3. In the early 2010s, ImmunoGen researchers began developing a drug to treat certain ovarian and peritoneal cancers. Specifically, ImmunoGen researchers tried to develop an immunoconjugate drug for patients with platinum-resistant ovarian cancer, which is especially difficult to treat because it does not respond to traditional cancer therapies. Most patients diagnosed with platinum-resistant ovarian cancer die within twelve months.

4. Immunoconjugates are a type of drug that treat cancers by specifically binding to a cell of interest and delivering a toxic drug payload to that cell. Immunoconjugates, which are also known as antibody drug conjugates or "ADCs," consist of three elements: (1) an antibody or the antigen-binding portion of an antibody which targets a specific cancer tumor antigen; (2) a toxic drug payload; and (3) a chemical linker that binds the antibody portion to the drug payload.

5. Immunoconjugates are considered large molecule drugs, in contrast to small molecule drugs such as antibiotics.¹

1 Small molecule drugs generally have a molecular weight between 150 and 280 grams per mole, while immunoconjugates have a molecular weight of about 150,000 grams per mole. See Tr. Nov. 1, 2022 (193:9-12) (Tolcher).

6. Immunoconjugates are difficult to develop because they often result in toxic side effects. For example, some immunoconjugates are known to cause ocular toxicity, which can result in symptoms such as dry eyes, blurred vision, and even blindness. Because of this, many immunoconjugates fail during the early stages of development.

7. Ocular toxicity is classified based on the National Cancer Institute Common Toxicity Criteria, which categorizes adverse ocular events from grades zero to five. Grade 0 indicates no toxicity; Grade 5 represents fatal toxicity. Grade 3 ocular toxicity is considered severe and can impair day-to-day life activities. In general, when patients experience ocular toxicity of Grade 3 and above during clinical trials, drug development is halted.

8. For over twenty years, ocular toxicity has been a well-known adverse event in the administration of immunoconjugates that contain as a toxic payload the maytansinoid² known as DM4.

2 Maytansinoids constitute a class of antitumor agent drugs derived from the chemical maytansine.

9. Dr. Lutz, a former ImmunoGen employee and one of the inventors of the '809 Application, testified that in the early 2010s, ImmunoGen originally tried to develop an immunoconjugate that contained the DM4 payload and an uncharged chemical linker. However, in both preclinical rabbit studies and human studies, that immunoconjugate caused severe ocular toxicity.

10. In response to the ocular toxicity that resulted from using an uncharged linker

with the DM4 payload, Dr. Lutz and other ImmunoGen researchers developed the immunoconjugate known as mirvetuximab soravtansine or IMGN853, which uses the DM4 payload with a charged linker instead of an uncharged one.

11. IMGN853 is thus comprised of (1) an antibody known as huMov19, (2) the DM4 payload, and (3) a charged chemical linker known as a charged sulfo-SPDB linker.³ IMGN853 functions as a cancer treatment because the antibody portion of IMGN853 targets ovarian and peritoneal cancer cells that express folate receptor 1 ("FOLR1"), and the toxic DM4 payload kills the cancer cells.

3 For this reason, IMGN853 is also sometimes referred to as "huMOV19-sulfo-SPDB-DM4."

12. Preclinical studies of IMGN853 in rabbits showed no ocular toxicity despite the use of the toxic DM4 payload. Because of this, Dr. Lutz believed that implementing the charged chemical linker had eliminated the previously observed ocular toxicity. Thereafter, Dr. Lutz applied for a patent for IMGN853 and published the results of the preclinical rabbit studies in Example 8 of the patent application known as Lutz '282.⁴ ImmunoGen ultimately received four patents covering IMGN853 and its use for treating cancer. See infra ¶¶ 102-07.

4 See U.S. Published Application No. 2012/0282282 (Lutz '282).

13. Based on the preclinical rabbit studies, Dr. Lutz and other ImmunoGen researchers proceeded in March 2012 to conduct human trials of IMGN853.

14. Even though the rabbit studies of IMGN853 reported no incidents of ocular toxicity, because the DM4 payload was known to cause adverse ocular side effects, ImmunoGen researchers continued to monitor for ocular toxicity during the clinical trials in humans.⁵ Furthermore, both Dr. Figg and Dr. Lutz testified that even though rabbits did not experience ocular toxicity when given IMGN853, preclinical results do not always translate to clinical safety when the same drug is tested in humans.⁶ Indeed, Dr. Lutz testified that he is familiar with instances where drugs tested in animals without any apparent toxicity later caused toxic side effects in humans.⁷

5 Tr. Nov. 1, 2022 (110:16-22) (Lutz).

6 Tr. Nov. 1, 2022 (113:15-16; 114:3-10) (Lutz); Tr. Nov. 3, 2022 (38:19-25; 39:1-12) (Figg).

7 Tr. Nov. 1, 2022 (114:3-10) (Lutz).

15. Thus, because IMGN853 includes the DM4 payload, a person of ordinary skill in the art would understand the potential risk of ocular toxicity and would monitor for ocular toxicity when testing IMGN853 in humans—just as Dr. Lutz testified that he did during the clinical trials of IMGN853.⁸

8 Tr. Nov. 1, 2022 (110:16-22) (Lutz).

16. ImmunoGen began the clinical trials of IMGN853 on March 1, 2012. During the initial Phase I trials in humans, IMGN853 was dosed according to the patient's total body weight ("TBW").

17. During the Phase I trials, ImmunoGen received confidential reports of ocular toxicity, just as a person of ordinary skill in the art would have expected as a possibility based on the known history of the DM4 payload, which IMGN853 contains.

18. Specifically, ImmunoGen received reports of Grade 3 adverse ocular events in patients who received a dose of IMGN853 at 7 milligrams per kilogram ("mg/kg") TBW. Three out of five patients who received IMGN853 at that dose experienced Grade 3 ocular toxicity. 19. After receiving reports of ocular toxicity in patients who received IMGN853 at a dose of 7 mg/kg TBW, ImmunoGen lowered the dose to 5 mg/kg TBW. Out of between ten to eleven patients⁹ who received the 5 mg/kg TBW dose, one patient experienced Grade 3 ocular toxicity.

9 Although the '809 Application reports that during the Phase I trials, ten patients received IMGN853 at a dose of 5 mg/kg TBW, ImmunoGen reported in later scientific publications that eleven patients received that dose.

20. Because one patient experienced Grade 3 ocular toxicity when dosed at 5 mg/kg TBW, ImmunoGen lowered the dose to 3.3 mg/kg TBW. At that dose, the drug did not result in sufficient therapeutic efficacy.

21. There is no data on 6 mg/kg TBW dosing of IMGN853, because during Phase I studies, ImmunoGen only tested doses of 3.3 mg/kg TBW, 5 mg/kg TBW, and 7 mg/kg TBW.

22. Thereafter, Drs. Ponte and Lutz, inventors of the '809 Application and then-employees of ImmunoGen, analyzed the pharmacokinetic ("PK") data collected during the Phase I trials. In doing so, Drs. Ponte and Lutz observed a correlation between early exposure to IMGN853 and ocular toxicity. They observed that, based on the Phase I data, ocular toxicity occurred when the early exposure PK parameters exceeded a certain threshold. They also observed that, based on the data from the Phase I trials, it appeared that appropriate levels of efficacy of IMGN853 could be achieved so long as the overall exposure to the drug reached a certain level.

23. Based on the PK data, the inventors sought a way to dose IMGN853 in order to reduce early exposure levels to stay below the ocular toxicity threshold while keeping the total overall exposure of IMGN853 above the threshold that correlated to therapeutic efficacy.

24. In looking for a manner of dosing IMGN853 to achieve a therapeutic window, Dr. Ponte found a formula for adjusted ideal body weight ("AIBW") on a nutrition website. Drs. Ponte and Lutz then conducted PK modeling of IMGN853 dosing regimens using the AIBW formula from the nutrition website. Drs. Ponte and Lutz also modeled dosing regimens using ideal body weight ("IBW") and body surface area.

25. Based on the PK modeling, Drs. Ponte and Lutz determined that AIBW dosing with the formula from the nutrition website was most likely to achieve a therapeutic window for IMGN853—that is, most likely to achieve therapeutically effective levels while remaining below the ocular toxicity threshold. Drs. Ponte and Lutz then decided to implement AIBW dosing in clinical trials of IMGN853.

26. On October 18, 2013, ImmunoGen amended its clinical trial protocol to change the dosing of IMGN853 from TBW to AIBW. Thereafter, ImmunoGen dosed IMGN853 in clinical trials at 5 mg/kg AIBW and 6 mg/kg AIBW. Seven patients received a dose of 5 mg/kg AIBW, and seven patients received a dose of 6 mg/kg AIBW. During these trials, the formula that ImmunoGen used to calculate AIBW was the same formula that Dr. Ponte found on the nutrition website.¹⁰ 27. None of the fourteen patients dosed at either 5 mg/kg AIBW or 6 mg/kg AIBW experienced Grade 3 or higher ocular events. At this point, none of the fourteen patients received prophylactic treatment such as eye drops.

10 Specifically, the AIBW equation from the nutrition website that Drs. Ponte and Lutz used to calculate AIBW during the clinical trials is as follows:

AIBW = IBW + 0.4 (TBW - IBW) Where IBW is calculated as below, with H indicating the patient's height in centimeters: IBW = 0.9H - 88 (males) IBW = 0.9H - 92 (females)

This is the same AIBW formula that is recited in Example 4 of the '809 Application. See infra ¶ 49.

28. In October 2014, the inventors filed the '809 Application currently at issue which seeks to patent the 6 mg/kg AIBW dosing regimen for IMGN853. The '809 Application is described in detail below.

C. The '809 Application

29. U.S. Patent Application No. 14/509,809, otherwise known as the '809 Application, is titled "Anti-FOLR1 Immunoconjugate Dosing Regimens" and was filed on October 8, 2014.

30. ImmunoGen is the assignee of the '809 Application.

31. The priority date for all claims of the '809 Application is October 8, 2013.

32. The '809 Application claims methods of treating patients with FOLR1-expressing ovarian or peritoneal cancers by administering the immunoconjugate known as mirvetuximab soravtansine or IMGN853.

33. Specifically, independent claims 1, 317, and 341 of the '809 Application read as follows:

1. A method for treating a human patient having an FOLR1-expressing cancer or cancer of the peritoneum comprising administering to the patient an immunoconjugate which binds to FOLR1 polypetide,

wherein the immunoconjugate comprises an antibody or antigen-binding fragment thereof that comprises the variable light chain (VL) complementarity determining region (CDR)-1, VH CDR-2, VL CDR-3, variable heavy chain (VH) CDR-1, VH CDR-2, and VH CDR-3 of SEQ ID Nos: 6-9, 11, and 12, respectively, and a maytansinoid, and

wherein the immunoconjugate is administered at a dose of 6 milligrams (mg) per kilogram (kg) of adjusted ideal body weight (AIBW) of the patient.

317. A method for treating a human patient having an FOLR1-expressing ovarian cancer comprising administering intravenously to the patient an immunoconjugate at a dose of 6 milligrams (mg) per kilogram (kg) of adjusted ideal body weight (AIBW) of the patient wherein the immunoconjugate comprises DM4 and an antibody comprising (i) a heavy chain comprising the same amino acid sequence as the amino acid sequence of the heavy chain encoded by the plasmid deposited with the American Type Culture Collection (ATCC) as PTA-10772 and (ii) a light chain comprising the same amino acid sequence as the amino acid sequence of the light chain encoded by the plasmid deposited with the ATCC as PTA-10774, and wherein the DM4 is linked to the antibody by sulfo-SPDB.

341. A method for treating a human patient having an FOLR1-expressing cancer of the peritoneum comprising administering intravenously to the patient an immunoconjugate at a dose of 6 milligrams (mg) per kilogram (kg) of adjusted ideal body weight (AIBW) of the patient wherein the immunoconjugate comprises DM4 and an antibody comprising (i) a heavy chain comprising the same amino acid sequence as the amino acid sequence of the heavy chain encoded by the plasmid deposited with the American Type Culture Collection (ATCC) as PTA-10772 and (ii) a light chain comprising the same amino acid sequence as the amino acid sequence of the light chain encoded by the plasmid deposited with the ATCC as PTA-10774, and wherein the DM4 is linked to the antibody by sulfo-SPDB.

34. The aforementioned claims 1, 317, and 341 are independent claims. The remaining

claims of the '809 Application are dependent claims that incorporate by reference all of the limitations recited in their respective independent claims. The dependent claims in the '809 Application recite a specific route of administration (i.e., claims 230, 259, and 260), a specific frequency of administration (i.e., claims 237, 238, 261, 262, 318, and 319), a specific cancer to be treated (i.e., claims 253-55, 258, 329, 330), a specific linker (i.e., claims 252, 300, 317, 341, 354), a specific maytansinoid (i.e., 242, 263-65, 320-25, 341, 349, and 354), or a specific antibody or antibody fragment (i.e., claims 266, 300, 317, 341, and 354).

35. The parties' Joint Submission of Stipulated Facts states that the only claims at issue in this litigation are claims 242, 252-55, 258-65, 300, 317-25, 329-30, 341-49, and 354 of the '809 Application.¹¹

11 See Joint Submission of Stipulated Facts, Dkt. 127 at ¶ 17.

36. IMGN853 is the commercial embodiment within the scope of the claims of the '809 Application.

37. Each of the independent claims of the '809 Application at issue in this litigation requires administering IMGN853 to patients at a dose of 6 mg/kg AIBW.

38. The immunoconjugate known as IMGN853 recited in the claims of the '809 Application was known in the art prior to October 8, 2013, as was a method of using IMGN853 to treat the cancers recited in claim 1 of the '809 Application. The '809 Application therefore seeks to patent only the method of dosing IMGN853 at 6 mg/kg AIBW.

D. Level of Ordinary Skill in the Art

39. At the time of the '809 Application, a person of ordinary skill in the art (a "POSA") would have an advanced degree in clinical pharmacy or clinical pharmacology and at least two years of experience in the field of clinical pharmacy, clinical pharmacology, or drug development with an emphasis on oncology agents, or would be a clinical oncologist with experience in drug development and similar years of experience.¹²

12 These Findings of Fact adopt the definition of a POSA provided by the PTO's expert witness, Dr. Figg. See Tr. Nov. 3, 2022 (13:9-15) (Figg). All three expert witnesses agreed, however, that their opinions would not change if they adopted the opposing side's view of the level of ordinary skill in the art. See Tr. Nov. 1, 2022 (159:24) (Tolcher); Tr. Nov. 2, 2022 (114:24-25; 115:1) (Shah); Tr. Nov. 3, 2022 (13:19) (Figg).

40. The three expert witnesses who testified during the bench trial—Drs. Tolcher, Shah, and Figg—each meet or exceed the definition a POSA. They are therefore in a position to render an opinion as to what a skilled artisan would have thought and understood regarding the issues relevant to this case.

41. Although each of the three expert witnesses qualifies as a POSA, Dr. Figg's testimony was the most credible and convincing. Dr. Figg's many titles support his credibility: he serves as an associate director of the Cancer Center at the National Cancer Institute, head of all clinical pharmacology for the National Institute of Health, and he oversees all drug development for the National Cancer Institute. Dr. Figg also serves as an adjunct professor of pharmacy at Virginia Commonwealth University and an adjunct professor of medical oncology at Columbia University. Dr. Figg has participated in drug development at the National Cancer Institute for 30 years and is experienced in clinical pharmacology and pharmacokinetics. Additionally, Dr. Figg has worked on developing hundreds of anti-cancer agents, has overseen drug development for ADCs, and has received 12

patents for various drugs and genetic tests.

E. Facts Pertaining to Indefiniteness

1. Intrinsic Record

42. Although the claims of the '809 Application are directed to administering IMGN853 at 6 mg/kg AIBW, none of the claims in the '809 Application specifies precisely how to calculate AIBW.

43. The "Definitions" section located in the specification of the '809 Application states that:

The term "adjusted ideal body weight" (AIBW) or "adjusted body weight" (ADJ) refers to a size descriptor that accounts for sex, total body weight, and height. AIBW and ADJ are used interchangeably throughout the specification. AIBW (ADJ) can be calculated, for example, using the formula ADJ = IBW + 0.4(weight in kg - IBW).

(emphasis added).

44. With respect to ideal body weight ("IBW"), which is used to calculate AIBW, the "Definitions" section states that:

The term "ideal body weight" (IBW) refers to a size descriptor that is unrelated to total body weight. IBW is an estimate of weight corrected for sex and height, and optionally frame size. IBW can be calculated, for example, using the formulas IBW = 0.9H-88 (for males) and IBW = 0.9H-92 (for females), wherein H = height in cm.

(emphasis added).

45. The "Definitions" section also states that IBW and AIBW "are discussed in more detail in Green and Duffull, British Journal of Clinical Pharmacology 58:119-133 (2004), which is herein incorporated by reference in its entirety."

46. The Green reference that the '809 Application incorporates in its entirety contains multiple formulas for calculating IBW and lists multiple different correction factors for AIBW.

47. Later, the '809 Application includes an "Examples" section, which contains a list of examples. The examples are prefaced with the following statement:

It is understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this application.

48. The conclusion of the "Examples" section also describes the examples as non-limiting, stating:

The breadth and scope of the present invention should not be limited by any of the above-described exemplary embodiments, but should be defined only in accordance with the following claims and their equivalents.

49. Example 4 of the '809 Application describes ImmunoGen's evaluation of dosing IMGN853 using IBW and AIBW. Example 4 defines IBW and AIBW in this context as follows:

Ideal Body Weight (IBW)

IBW (male) = 0.9H-88

IBW (fem) = 0.9H-92

(where H = height in cm)

...

Adjusted Ideal Body Weight (AIBW or ADJ)

IBW + 0.4(Actual weight in kg - IBW)

50. The PTO did not reject the claims of the '809 Application as indefinite during prosecution, and instead first raised indefiniteness in this civil action.

51. During the patent prosecution process of the '809 Application, the PTAB stated that "[ImmunoGen] defines '[t]he term "ideal body weight" (IBW) ... [as] a

size descriptor that is unrelated to total body weight. IBW is an estimate of weight corrected for sex and height, and optionally frame size. IBW can be calculated, for example, using the formulas IBW = 0.9H-88 (for males) and IBW = 0.9H-92 (for females), wherein H = height in cm.'" (emphasis added).

52. The PTAB also stated that "[ImmunoGen] defines '[t]he term "adjusted ideal body weight" (AIBW) or "adjusted body weight" (ADJ) ... [as] a size descriptor that accounts for sex, total body weight, and height ... AIBW (ADJ) can be calculated, for example, using the formula ADJ = IBW + 0.4(weight in kg - IBW).'" (emphasis added).

53. As Dr. Figg credibly testified, the claim language in the '809 Application would leave a POSA "confused over how to calculate [AIBW]" because the Application's inclusion of the term "for example" in the "Definitions" section leaves the definition of AIBW "vague."¹³

13 Tr. Nov. 3, 2022 (24:21-25; 25:19-20) (Figg).

54. Nor would a POSA understand based solely on the terms of the '809 Application how to calculate IBW, which is a necessary factor in calculating AIBW. Dr. Figg credibly testified that a POSA reading the '809 Application would not know how to calculate IBW because the application's "Definitions" section expressly states that IBW can include an "optional[]" frame size correction and twice uses the term "for example."¹⁴ Dr. Figg also testified that the '809 Application's incorporation of the Green reference would leave a POSA "unclear as to which equation to use," because Green lists various different methods for calculating AIBW and IBW.¹⁵

14 Tr. Nov. 3, 2022 (25:21-25; 26:1-3) (Figg).

15 Tr. Nov. 3, 2022 (26:11-15) (Figg).

2. Extrinsic Record

55. Evidence extrinsic to the '809 Application and prosecution history points to multiple methods of calculating AIBW. As of the '809 Application's priority date of October 8, 2013, there were many ways to define AIBW, all of which involved a variation of the following generic formula, wherein "CF" stands for "correction factor":

AIBW = IBW + CF (weight in kg - IBW).

56. The art prior to October 8, 2013 disclosed a wide range of AIBW correction factors used for different drugs. Dr. Figg testified that although 0.4 is often used as a standard AIBW correction factor, correction factors for use in AIBW formulas can range from 0.2 to 0.9.¹⁶ And ImmunoGen's expert, Dr. Shah, admitted that the choice in correction factor can lead to a "huge" difference in the resulting dose.¹⁷

16 Tr. Nov. 3, 2022 (15:11-13) (Figg).

17 Tr. Nov. 2, 2022 (143:8-12) (Shah).

57. Furthermore, some drugs have more than one reported AIBW correction factor. For example, there are at least three correction factors reported for the drug tobramycin, and two different correction factors reported for the drugs gentamicin and amikacin.

58. The extrinsic record also makes clear that there are numerous ways to calculate IBW, which is necessary to calculate for use in the AIBW formula. As of the priority date of the '809 Application, there were at least eleven formulas for calculating IBW.

59. This extrinsic evidence further leaves a POSA uncertain as to the scope of the claims in the '809 Application. Dr. Figg credibly testified that the numerous different

correction factors reported in the literature would leave a POSA uncertain as to which correction factor to use when calculating AIBW for IMGN853 pursuant to the '809 Application. And Dr. Figg also testified that a POSA reading the '809 Application would not know how to calculate IBW, given that the '809 Application only provides a specific IBW formula as an "example" and there are at least eleven possible IBW formulas in the extrinsic record.¹⁸

18 Tr. Nov. 3, 2022 (15:20-25; 22:24-25; 23:2-7) (Figg).

60. Based on the extrinsic evidence, a POSA would not understand that the formula for AIBW should be limited to the formula listed in Example 4 of the '809 Application. Although ImmunoGen's expert, Dr. Shah, testified that a POSA would understand that the formula for AIBW was limited to the one listed in Example 4 of the '809 Application,¹⁹ Dr. Figg credibly testified that a POSA would not read Example 4 as limiting the definition of AIBW. This is because, as Dr. Figg explained, "the literature is filled with ... different equations for ideal body weight and a lot of different correction factors."²⁰ Furthermore, as Dr. Figg explained, Example 4 is prefaced by language stating that the examples in the '809 Application are for illustrative purposes only.²¹

19 Tr. Nov. 2, 2022 (115:4, 10-13) (Shah).

20 Tr. Nov. 3, 2022 (27:9-11) (Figg).

21 Tr. Nov. 3, 2022 (27:21-22) (Figg).

F. Facts Pertaining to Obviousness

1. Scope and Content of the Prior Art

61. The drug IMGN853, for which ImmunoGen owns multiple patents, is an embodiment of the immunoconjugate recited in the '809 Application. The parties do not dispute that IMGN853 was known in the art prior to October 8, 2013. IMGN853 was disclosed in the following prior art references, each of which is described in further detail below: (1) U.S. Published Application No. 2012/0282282 ("Lutz '282"); (2) International Published Application No. WO 2011/106528 ("Ab '528"); (3) U.S. Published Application No. 2012/0009181 ("Ab '181"); (4) International Published Application No. WO 2012/135675 ("Carrigan '675"); (5) U.S. Published Application No. 2012/0282175 ("Carrigan '175"); (6) International Published Application No. WO 2012/138749; and (7) a 2013 abstract presented at the American Society of Clinical Oncology (the "2013 ASCO abstract"). Additionally, ImmunoGen ultimately received four patents covering IMGN853 and its use for treating cancer. See infra ¶¶ 102-07.

a. Lutz '282, a patent application for IMGN853 filed in 2012 by Dr. Lutz, discloses that IMGN853 could be used to treat FOLR1-expressing ovarian or peritoneal cancer.

b. Lutz '282 reports that immunoconjugates like IMGN853 are known to cause ocular toxicity. Lutz '282 reports that using a charged chemical linker reduced ocular toxicity in rabbits, but does not report any results of human testing.

c. Lutz '282 focuses on dosing IMGN853 around 6 mg/kg TBW. Specifically, Lutz '282 states the following:

In one embodiment, the dose of ADC administered is at least about 4 mg/kg. In another embodiment, the dose is between about 4 mg/kg and about 16 mg/kg. In another embodiment, the dose is between about 4 mg/kg and about 8 mg/kg. In another embodiment, the dose is between about 5 mg/kg and 6 mg/kg. In another embodiment, the dose is between about 6 mg/kg and about 8 mg/kg. In a further embodiment, the dose is between about 6 mg/kg and about 7 mg/kg. In another embodiment, the dose is between about 7

mg/kg and about 8 mg/kg. In yet another embodiment, the dose is between about 4 mg/kg and 6 mg/kg. In a further embodiment, the dose is between about 4 mg/kg and 5 mg/kg.

(emphasis added).

d. Ab '528 describes another IMGN853 study by ImmunoGen and discloses that IMGN853 is a promising treatment for ovarian and peritoneal cancers based on mouse tumor models. In terms of dosing IMGN853, Ab '528 discloses that the "administering physician can easily determine optimum dosages, dosing methodologies and repetition rates."

e. Ab '181 similarly discloses that IMGN853 shows promising activity in mouse tumor models. ImmunoGen has received U.S. Patent No. 8,557,966 based on the application published as Ab '181.

f. Carrigan '675 similarly describes IMGN853 as an ovarian cancer treatment and states that an "administering physician can easily determine optimum dosages, dosing methodologies and repetition rates."

g. Carrigan '175 also discloses IMGN853.

h. International Published Application No. WO 2012/138749 also discloses IMGN853.

i. The 2013 ASCO abstract was presented by ImmunoGen researchers at a conference in May-June 2013. The abstract describes ImmunoGen's preliminary results from Phase I testing of IMGN853 in human patients dosed based on TBW. The abstract describes that one patient with serious ovarian cancer had an 82% reduction in tumor marker CA125 after receiving IMGN853 at a dose of 3.3 mg/kg TBW. The 2013 ASCO abstract also discloses that patients received IMGN853 at doses up to 5 mg/kg TBW and concludes that "[s]afety evaluation continues at 5 mg/kg [TBW] and dose escalation [of IMGN853] is ongoing."

62. AIBW was also well known in the prior art as a possible dosing mechanism for various treatments, including for multiple antibiotics and at least six anticancer treatments. Indeed, Dr. Figg catalogued eleven separate prior art references that teach using AIBW dosing, each of which is described below.

a. First, Dr. Figg's group published an article in 2003 which concludes that AIBW dosing decreased the variability in the clearance²² of the anticancer drug known as COL-3.²³

22 "Clearance" is a pharmacokinetic term describing the rate at which the body eliminates drug molecules.

23 See Michelle A. Rudek et al., Factors Involved in the Pharmacokinetics of COL-3, a Matrix Metalloproteinase Inhibitor, in Patients with Refractory Metastatic Cancer: Clinical and Experimental Studies, 43 J. Clinical Pharmacology 1124 (2003).

b. Fuchs et al. filed a patent application prior to October 2013 that teaches using AIBW to dose cyclophosphamide, an anticancer agent.²⁴

24 See U.S. Published Application No. 2012/0148577.

c. Geraghty et al. also filed a patent application prior to the '809 Application's priority date which discloses using AIBW to dose thiotepa, another anticancer agent.²⁵

25 See U.S. Published Application No. 2009/0162374.

d. An article authored by Ekhart et al. teaches using AIBW for administering the anticancer drug carboplatin.²⁶

26 See Corine Ekhart et al., Carboplatin Dosing in Overweight and Obese Patients with Normal Renal Functions, Does Weight Matter?, 64 Cancer Chemotherapy & Pharmacology 115 (2009).

e. Like the Ekhart reference, an article authored by Herrington et al. also teaches

using AIBW for administering the anticancer drug carboplatin.²⁷

27 See Jon D. Herrington et al., Prospective Evaluation of Carboplatin AUC Dosing in Patients with a BMI =27 or Cachexia, 57 Cancer Chemotherapy & Pharmacology 241 (2006).

f. A reference authored by Topcuoglu et al. teaches using AIBW to determine how many CD34 + cells to transplant for treating certain cancers.²⁸

28 See Pervin Topcuoglu et al., How to Calculate the Quantity of CD34 + Cells Infused? A Single Center Cohort Study Based on Actual, Ideal or Adjusted Body Weight, 36 Transfusion and Apheresis Science 275 (2007).

g. Like the Topcuoglu reference, an article authored by Hicks et al. teaches using AIBW to determine how many CD34 + cells to transplant for treating certain cancers.²⁹

29 See C. Hicks et al., The Use of Adjusted Ideal Body Weight for Overweight Patients Undergoing HPC Mobilisation for Autologous Transplantation, 91 Annals of Hematology 1795 (2012).

h. The FDA-approved package insert included with the anticancer chemotherapy drug known as Busulfex recommends using AIBW dosing for obese patients.

i. Green, a reference published in 2004 and incorporated in its entirety in the '809 Application, teaches using AIBW for dosing various antibiotics and provides various formulas for calculating IBW and AIBW.³⁰

30 See Bruce Green & Stephen B. Duffull, What is the Best Size Descriptor to Use for Pharmacokinetic Studies in the Obese?, 58 British J. of Clinical Pharmacology 119 (2004).

j. Similarly, Bauer, a reference published in 1983, discusses AIBW as a method for dosing various antibiotics.³¹

31 See L.A. Bauer et al., Influence of Weight on Aminoglycoside Pharmacokinetics in Normal Weight and Morbidly Obese Patients, 24 European J. of Clinical Pharmacology 643 (1983).

k. Finally, an article authored by Alffenaar & van der Werk describes ocular toxicity caused by an antibiotic known as ethambutol. The article states that the researchers considered using AIBW dosing for ethambutol to ameliorate ocular toxicity, but ultimately settled on IBW dosing.³²

32 See Jan-Willem Alffenaar & Tjip van der Werf, Dosing Ethambutol in Obese Patients, 54 Antimicrobial Agents & Chemotherapy 4044 (2010).

63. Based on these prior art references, a POSA as of October 2013 would have been familiar with AIBW as a dosing method used for various anticancer drugs and for at least one cell transplant therapy used to treat cancer patients. A POSA would therefore have been familiar with AIBW as a potential dosing method for immunoconjugates like IMGN853.

64. The specific AIBW formula listed in the "Definitions" section and Example 4 of the '809 Application was also known in the prior art. Drs. Lutz and Ponte, two of the inventors of the '809 Application, testified that they found the AIBW formula recited in Example 4 on a nutritional website.³³ Because of this, ImmunoGen's expert Dr. Tolcher testified that "the discovery" claimed in the '809 Application "is not the [AIBW] formula," which was found on a nutritional website.³⁴

33 Tr. Nov. 1, 2022 (73:5) (Ponte); Tr. Nov. 1, 2022 (102:21-25) (Lutz).

34 Tr. Nov. 1, 2022 (178:8-13) (Tolcher).

65. Furthermore, the 0.4 correction factor used in the '809 Application's AIBW formula was not only disclosed in the prior art, but was also commonly used as a standard correction factor for AIBW dosing since the 1980s.³⁵ 2. Differences Between the Prior Art and the Claims at Issue

35 Tr. Nov. 3, 2022 (46:18-19; 103:20-24) (Figg).

66. Although Lutz '282 focuses on dosing IMGN853 at around 6 mg/kg TBW, no prior art reference discloses dosing IMGN853 at 6 mg/kg AIBW.

67. No prior art reference describes using AIBW to dose an immunoconjugate drug.

68. Most of the prior art references describe AIBW dosing for small molecule drugs, such as carboplatin, cyclophosphamide, and thiotepa. Two references, however, disclose AIBW dosing for a large molecule treatment. Topcuoglu and Hicks et al. both discuss AIBW dosing for CD34 + transplant cells, which are thousands of times larger than immunoconjugates.

69. The prior art focuses on using AIBW dosing in obese or overweight patients. However, the article published in 2003 by Dr. Figg's group discusses using AIBW to dose COL-3 to treat metastatic cancer for every type of patient, independent of weight.³⁶

36 See Rudek et al., supra note 23.

70. Although it was well known prior to October 2013 that the DM4 payload in immunoconjugates caused ocular toxicity, no prior art reference specifically disclosed that IMGN853 caused ocular toxicity in humans.

71. The prior art does not disclose the pharmacokinetic data described in paragraphs [00195] through [00205] or Figures 1B, 2A-C, and 3-8 of the specification of the '809 Application. Nor does the prior art disclose pharmacokinetic data for IMGN853 identifying a correlation between early exposure and ocular toxicity.

72. No prior art reference recommends using AIBW dosing in order to reduce ocular toxicity. Although Alffenaar & van der Werk's reference describes that researchers considered using AIBW to dose ethambutol in order to reduce ocular toxicity, the researchers ultimately recommend IBW dosing instead of AIBW dosing.

3. Motivation to Combine

73. A POSA as of October 8, 2013 would have been motivated to combine the prior art references and try AIBW dosing as a potential dosing methodology for IMGN853 in order to eliminate ocular toxicity. This is because AIBW was a well-known dosing methodology, had been used in both small and large molecule drugs, and had even been used to dose various cancer drugs.

74. Even though no prior art reference disclosed ocular toxicity in humans attributable to the administration of IMGN853, a POSA would nevertheless understand the potential risk of ocular toxicity from a drug that included the DM4 payload and would monitor for ocular toxicity when testing IMGN853 in humans, just as Dr. Lutz testified that he did during the clinical trials of IMGN853.³⁷

37 Tr. Nov. 1, 2022 (110:16-22) (Lutz).

75. Once a POSA observed ocular toxicity from the administration of IMGN853 in the ordinary course of human trials, it would be obvious to the skilled artisan to experiment with changing the dosing methodology to reduce toxicity while preserving the therapeutic effects. Indeed, Dr. Figg testified that when confronted with toxicities, persons of ordinary skill in the art will generally consider changing the dose.³⁸ And even one of ImmunoGen's experts admitted that although the "pathophysiology of the ocular toxicity" is unknown, the toxicity associated with the DM4 payload "is usually dose-related."³⁹ 76. In experimenting with dosing methodologies for IMGN853, a POSA would try AIBW dosing, given that AIBW dosing was well known in the prior art as a dosing strategy, including as a strategy for dosing anticancer drugs. Indeed, as Dr. Figg testified, "it is common knowledge to [use AIBW] in the field of drug development of anticancer agents."⁴⁰

38 Tr. Nov. 3, 2022 (56:19-23) (Figg).

39 Tr. Nov. 1, 2022 (164:23-25; 165:5-8) (Tolcher).

40 Tr. Nov. 3, 2022 (41:19-20) (Figg).

77. A POSA seeking a dosing methodology for IMGN853 would look to prior art references pertaining to both large and small molecule drugs. Even though immunoconjugates are large molecule drugs, Dr. Figg explained that medical oncologists are familiar with both small and large molecule drugs for use in treating cancer. Thus, the prior art references discussing the use of AIBW in small molecule drugs would motivate a POSA to experiment with AIBW dosing for IMGN853.⁴¹

41 Tr. Nov. 3, 2022 (43:9-13; 45:11-24) (Figg).

78. Additionally, a POSA would be motivated to try AIBW dosing in order to reduce the ocular toxicity associated with IMGN853 because Alffenaar & van der Werk's reference describes that before October 2013, researchers had already considered using AIBW to reduce ocular toxicity in patients receiving the drug ethambutol.

79. Indeed, around the same time as Drs. Lutz and Ponte were developing the '809 Application and before October 2013, other researchers were motivated to try AIBW dosing in order to reduce toxic side effects in a clinical study on a radioimmunoconjugate⁴² known as actinium-225-lintuzumab. From February 2006 to March 2012, researchers tested the radioimmunoconjugate and originally dosed the drug using TBW.⁴³ After the researchers observed toxic side effects, however, the researchers decided to try AIBW dosing for patients above their ideal body weight. This study was not reported until 2022, so it is not prior art for the purpose of the '809 Application.⁴⁴ However, Rosenblat et al. demonstrates that a POSA as of October 2013 would have been motivated to experiment with AIBW dosing, as it shows that other researchers were already doing so before the priority date of the '809 Application.

42 Radioimmunoconjugates are similar to immunoconjugates, but they contain a radiolabel that kills cancer cells instead of a toxic drug payload. Both radioimmunoconjugates and immunoconjugates are large molecule drugs.

43 See Todd L. Rosenblat et al., Treatment of Patients with Acute Myeloid Leukemia with the Targeted Alpha-Particle Nanogenerator Actinium-225-Lintuzumab, 28 Clinical Cancer Research 2030 (2022).

44 Although the Rosenblat reference is not part of the prior art in this case because it was not reported until 2022, the Federal Circuit has instructed that "independently made, simultaneous inventions" made "within a comparatively short space of time" of the claimed invention can provide "persuasive evidence that the claimed [invention] was the product only of ordinary mechanical or engineering skill." Geo. M. Martin Co. v. Alliance Mach. Sys. Int'l LLC, 618 F.3d 1294, 1305 (Fed. Cir. 2010) (quoting Concrete Appliances Co. v. Gomery, 269 U.S. 177, 184, 46 S.Ct. 42, 70 L.Ed. 222 (1925)).

80. In 2012, ASCO convened a panel that recommended using TBW, not AIBW, to dose chemotherapy agents (the "2012 ASCO panel"). For multiple reasons, this recommendation would not have dissuaded a POSA from experimenting with AIBW dosing for IMGN853. First, the FDA did not change its requirement of using AIBW dosing in obese patients for the chemotherapy drug Busulfex after the 2012 ASCO panel. Second, Dr. Figg testified that the 2012 ASCO panel would not have

dissuaded an oncologist from using AIBW dosing when treating patients.⁴⁵ Indeed, even Dr. Tolcher testified that he "can't disagree" with the statement that the ASCO 2012 recommendation not to use AIBW dosing "isn't very persuasive."⁴⁶ Finally, the 2012 ASCO panel only recommended TBW for chemotherapy drugs, and IMGN853 is not a chemotherapy drug.

45 Tr. Nov. 3, 2022 (54:14-19) (Figg).

46 Tr. Nov. 1, 2022 (189:14-17) (Tolcher).

81. A POSA seeking to experiment with AIBW dosing for IMGN853 would start with dosing around 6 mg/kg AIBW, because the prior art focuses on doses of about 5 to 6 mg/kg TBW. Lutz '282 focused on doses around 6 mg/kg TBW,⁴⁷ and the 2013 ASCO Abstract instructed that doses of 5 mg/kg TBW were being tested. Thus, a skilled artisan seeking to determine the proper AIBW dose would start with doses of either 5 or 6 mg/kg AIBW and determine the best dose through routine optimization—just as ImmunoGen's researchers did.

47 See Tr. Nov. 3, 2022 (33:9-16) (Figg).

82. A POSA experimenting with AIBW dosing for IMGN853 would also be motivated to try the 0.4 correction factor, because as Dr. Figg testified, 0.4 has been known as a standard AIBW correction factor since the 1980s. Indeed, Dr. Figg testified that researchers often start with the 0.4 correction factor when experimenting with AIBW dosing.⁴⁸

48 Tr. Nov. 3, 2022 (103:20-24) (Figg).

4. Reasonable Expectation of Success

83. Even though developing immunoconjugates for treating certain cancers is challenging due to potential toxic side effects, a POSA would have a reasonable expectation of success in experimenting with AIBW dosing for immunoconjugates in order to reduce ocular toxicity. Although, as Dr. Figg testified, the exact mechanism by which immunoconjugates cause ocular toxicity is "not completely known,"⁴⁹ it was well understood that the DM4 payload can result in ocular toxicity. Furthermore, as Dr. Figg explained, there was no question that dose was relevant to toxicity. Thus, Dr. Figg testified that a POSA would attempt to change IMGN853's dose when confronted with ocular toxicity because it was well understood by persons of ordinary skill in the art that "[i]f you give too much, you get toxicity."⁵⁰

49 Tr. Nov. 3, 2022 (111:7-9) (Figg).

50 Tr. Nov. 3, 2022 (56:3) (Figg).

84. In deciding how to alter the dose for IMGN853, a POSA would have a finite number of options to try: Dr. Figg testified that the standard methods of altering the dosing for anticancer drugs in order to reduce side effects are (1) dose reduction, (2) changing the dose frequency, or (3) assessing how to calculate the dose.⁵¹ In terms of assessing how to calculate the dose, Dr. Figg testified that there are four options other than using total body weight, all of which are disclosed in the Green reference: (1) body surface area, (2) ideal body weight, (3) lean body weight, and (4) AIBW dosing.⁵² Indeed, ImmunoGen's witness, Dr. Ponte, similarly testified that in considering alternative metrics for calculating body weight when developing the '809 Application, he considered body surface area, ideal body weight, lean body weight, and AIBW.⁵³

51 Tr. Nov. 3, 2022 (40:5-7) (Figg).

52 Tr. Nov. 3, 2022 (40:12-13) (Figg).

53 Tr. Nov. 1, 2022 (60:23-25; 61:1-3) (Ponte).

85. A POSA would also have a reasonable expectation of success in experimenting

with AIBW dosing based on the Alffenaar & van der Werk reference, which was in the prior art as of October 2013. That article states that researchers considered using AIBW dosing in order to reduce ocular toxicity associated with the antibiotic ethambutol.

86. Finally, although not part of the prior art because it was not reported until 2022, research on the radioimmunoconjugate actinium-225-lintuzumab further supports that a POSA would have a reasonable expectation of success in experimenting with AIBW dosing for IMGN853. The Rosenblat et al. reference demonstrates that that other researchers were contemporaneously experimenting with AIBW dosing in order to reduce ocular toxicity around the same time as ImmunoGen researchers were developing IMGN853 and the '809 Application.

5. Secondary Indicia of Obviousness

i. Unexpected Results

87. ImmunoGen claims that dosing IMGN853 at 6 mg/kg AIBW unexpectedly resulted in the elimination of ocular toxicity for patients across the weight spectrum. In support, ImmunoGen points to the early Phase I results which disclosed no incidents of ocular toxicity in the small group of individuals initially given IMGN853 dosed based on AIBW. But more recent studies of AIBW dosing for IMGN853 call into question the early Phase I results and do not in fact confirm that AIBW dosing eliminates or significantly reduces incidents of ocular toxicity. Thus, as explained below, there is insufficient information based on the current record to determine whether 6 mg/kg AIBW dosing does in fact reduce ocular toxicity for IMGN853.⁵⁴

54 At the time of the bench trial, the FDA had not granted approval for IMGN853. Shortly after the trial was completed, however, the FDA granted accelerated approval, or "Fast Track" approval, for IMGN853 to be administered to patients. See https://www.accessdata.fda.gov/drugsatfda_docs/appletter/2022/761310Orig1s000ltr.pdf (last accessed Jan. 30, 2023) (hereinafter "FDA Approval Letter"). As part of that approval, the FDA stated that ImmunoGen is required to "[c]onduct a randomized clinical trial of at least 2 dose schedules of mirvetuximab soravtansine to characterize the safety of the recommended dosage of 6 mg/kg AIBW Q3W and alternative dosing schedule(s) (i.e., modified weekly dosing regimen, where lower doses are administered at a weekly frequency, instead of every 3 weeks.)" Id. at 5. Thus, at this time, ImmunoGen has not carried its burden of showing that AIBW dosing for IMGN853 does in fact reduce ocular toxicity.

88. First, there is only a small amount of data for patients dosed based on TBW. In a Phase I study, three out of five patients who received IMGN853 at a dose of 7 mg/kg TBW experienced Grade 3 ocular toxicity. Out of the ten to eleven patients who received IMGN853 at a dose of 5 mg/kg TBW, one patient experienced Grade 3 ocular toxicity.

89. After the researchers decided to switch to AIBW in the Phase I study, seven patients received a dose of 6 mg/kg AIBW. None of those patients experienced Grade 3 ocular toxicity. At that point, ImmunoGen did not give any of the patients prophylactic measures such as lubricating eye drops.

90. The Phase I results reporting zero incidences of Grade 3 ocular toxicity in patients who received 6 mg/kg AIBW were not confirmed in later, larger studies of IMGN853. As described below, when ImmunoGen administered the 6 mg/kg AIBW dose to larger cohorts during two later studies called the "Forward 1" and "Soraya" studies, there were incidents of Grade 3 ocular toxicities, even after the patients were given numerous prophylactic treatments such as lubricating eyedrops to minimize ocular toxicity. 91. First, the Forward 1 study reported the following Grade 3 ocular toxicities: 2% blurred vision, 1% keratopathy, and 1% dry eye. Dr. Figg explained that during the course of the Forward 1 study, ImmunoGen made important adjustments to its protocols to reduce the frequency and severity of ocular toxicity.⁵⁵ Namely, after ImmunoGen observed Grade 2 blurred vision in roughly 54% of the patients, ImmunoGen implemented procedures including excluding patients from the study with active or chronic ocular conditions, advising patients experiencing eye problems to apply a warm compress at bedtime, and recommending the daytime use of sunglasses and lubricating eye drops. After ImmunoGen implemented these procedures, the percentage of patients with blurred vision dropped to around 39% and was mostly Grade 1 ocular toxicity.

55 Tr. Nov. 3, 2022 (72:16-74:9) (Figg).

92. Because of these measures—which were not implemented during the Phase I study of IMGN853—Dr. Figg explained that the ocular toxicity numbers in the Forward 1 study are "falsely lower" than they would have been had ImmunoGen applied the same protocols that it did during the Phase I studies.⁵⁶ And despite all of the added prophylactic measures used in Forward 1, Dr. Figg testified that ImmunoGen's more recent publications show that 20% of patients in the Forward 1 study needed either delayed dosing or dose reductions due to ocular toxicity, confirming that ocular toxicity remains a serious problem for IMGN853 even with AIBW dosing.⁵⁷

56 Tr. Nov. 3, 2022 (69:8) (Figg).

57 Tr. Nov. 3, 2022 (77:11-25) (Figg).

93. Second, the Soraya study included all of the same prophylactic measures designed to reduce ocular toxicity as the Forward 1 study. Even with all of the additional protocols, 12 patients, or 11%, experienced Grade 3 ocular toxicity. Thus, the Soraya study does not confirm the Phase I results that AIBW dosing unexpectedly eliminates ocular toxicity in patients receiving IMGN853.

94. According to ImmunoGen, a poster from a September 2022 meeting of the European Society of Clinical Oncology ("ESMO") confirms that AIBW dosing reduced Grade 3 ocular toxicities in patients across the weight spectrum. The ESMO poster reports the results of a covariate analysis conducted using the data from ImmunoGen's Phase I studies, the Forward I study, and the Soraya study. The ESMO poster reports that AIBW was not a covariate that impacted the efficacy of IMGN853 and its dosing regimen. Dr. Shah testified that this covariate analysis means that for all patients, regardless of weight, dosing IMGN853 at 6 mg/kg AIBW results in lower incidences of toxicity. In other words, according to Dr. Shah, underweight individuals dosed with 6 mg/kg AIBW are not "getting more toxicity than overweight individual[s]."⁵⁸ On the other hand, Dr. Figg testified that the results reported in the ESMO poster indicate that "AIBW had limited impact on efficacy and safety, and [is] likely not of clinical significance."⁵⁹ Because the experts in this case had little time to review the ESMO poster,⁶⁰ very little testimony was presented about the methodology or reliability of covariate studies, and no evidence was presented about the underlying assumptions in this covariate analysis, the

58 Tr. Nov. 3, 2022 (170:23-25) (Shah).

59 Tr. Nov. 3, 2022 (83:16-18) (Figg).

60 The ESMO poster was only recently published in September 2022 and was not disclosed by ImmunoGen in the ordinary course of discovery. See Order Issued Nov. 3, 2022, Dkt. 136, at 1-2.

ESMO poster and its conclusions are afforded little weight here.

95. On this record, there is not enough information to determine whether AIBW dosing is safe and effective for patients of all weights. In fact, as Dr. Figg testified, AIBW dosing may be dangerous for underweight patients.⁶¹ This is because for patients who weigh less than their ideal body weight, AIBW dosing results in the patients receiving a higher dose of IMGN853 than the patients would otherwise receive if they were dosed based on TBW. In contrast, patients who are overweight or obese receive a lower amount of IMGN853 when dosed according to AIBW than they would receive if dosed according to TBW; patients who weigh exactly their ideal body weight receive an identical dose of IMGN853 when dosed based on either AIBW or TBW.

61 Tr. Nov. 3, 2022 (129:2-3) (Figg).

96. As Dr. Figg explained, cachectic or underweight patients receive a larger amount of the toxic DM4 payload under the claimed AIBW dose for IMGN853 than they otherwise would if they were dosed based on TBW.⁶² According to Dr. Figg, under the claimed AIBW dosing regimen, underweight patients may receive as much as 30 percent more drug than they would receive if dosed based on TBW.⁶³ Because of this, Dr. Figg testified that the persisting incidents of ocular toxicity in the Forward 1 and Soraya studies may have occurred in underweight patients who received too much of the toxic payload.⁶⁴ Of course, further studies in the future may reveal that AIBW dosing is indeed therapeutically effective for patients across the weight spectrum, but based on the evidence presented at trial, there is insufficient information upon which to draw this conclusion.

62 Tr. Nov. 3, 2022 (129:5-8) (Figg).

63 Id.

64 Tr. Nov. 3, 2022 (129:8-11; 15-17) (Figg).

ii. Long-felt Need

97. The '809 Application is directed to a method of administering IMGN853 in order to reduce ocular toxicity. The ocular toxicity problems associated with the administration of IMGN853 were first observed during Phase I testing in 2012, and were not publicly disclosed until 2017. Thus, there was no publicly recognized, long-felt need for AIBW dosing for IMGN853 as of October 8, 2013.

iii. Failure of Others

98. Other drug companies, including Sanofi, Bayer, AbbVie, and Seattle Genetics, have failed to develop immunoconjugates because they were not able to overcome toxic side effects.

iv. Copying

99. After the filing of the '809 Application, one of ImmunoGen's competitors called Sutro Biopharma, Inc. adopted AIBW dosing in clinical trials for an immunoconjugate that, like IMGN853, targets FOLR1 to treat ovarian cancer.

v. Industry Praise

100.vImmunoGen introduced two documents that praised IMGN853: a trade publication and an investment analyst sheet from Canaccord Genuity.⁶⁵ Although these documents praise the anticancer properties of IMGN853, neither praises AIBW dosing

65 The parties also agreed in their Joint Submission of Stipulated Facts that an article coauthored by Dr. Figg stated that IMGN853 "has shown promising activity in heavily treated ovarian cancer patients" and "demonstrated a favorable benefit-risk profile." Dkt. 127, ¶ 24. Neither party, however, introduced the article into evidence.

as a method for reducing ocular toxicity. And a trade publication and investment analyst sheet are not particularly impressive sources of praise as would be, for example, statements by experts in the field.

101. And although Dr. Figg himself praised IMGN853, emphasizing that he "hope[s] this drug gets approved" because "[w]e need more drugs for ovarian cancer," Dr. Figg made clear that none of the cited industry praise is "talking about ... AIBW."⁶⁶

66 Tr. Nov. 3, 2022 (85:1-5) (Figg).

G. Facts Pertaining to Obviousness-Type Double Patenting

102. ImmunoGen owns the '809 Application as well as U.S. Patent Nos. 8,557,966 (the "'966 patent"); 9,133,275 (the "'275 patent"); 9,598,490 (the "'490 patent); and 9,670,280 (the "'280 patent").

103. The '966 patent was granted on October 15, 2013. Claim 1 of ImmunoGen's '966 patent recites a specific kind of "humanized antibody or antigen binding fragment thereof," and IMGN853 is an embodiment of that antibody or fragment.

104. The '275 patent was granted on September 15, 2015. Claim 1 of ImmunoGen's '275 patent recites a "method of treating cancer" by "administering a therapeutically effective amount" of a specific kind of antibody, and IMGN853 is an embodiment of that antibody.

105. The '490 patent was granted on March 21, 2017. Claim 1 of ImmunoGen's '490 patent recites a specific immunoconjugate, and IMGN853 is an embodiment of that immunoconjugate.

106. The '280 patent was granted on June 6, 2017. Claim 1 of ImmunoGen's '280 patent recites a "method of treating cancer" by "administering a therapeutically effective amount" of a specific kind of antibody, and IMGN853 is an embodiment of that antibody.

107. The '966, '275, '490, and '280 patents indicate that ImmunoGen has already received multiple patents for IMGN853 and its use for treating cancer.⁶⁷

67 Because of this, any entity other than ImmunoGen that attempted to find the proper dose for IMGN853 to preserve its therapeutic effect while reducing ocular toxicity would be infringing on ImmunoGen's patents. And even if an exception for experimental use applied allowing another entity to experiment with dosing for IMGN853, no entity other than ImmunoGen would have an incentive to do so, as only ImmunoGen could profit from use of the drug.

III. CONCLUSIONS OF LAW

Set forth below, pursuant to Rule 52, Fed. R. Civ. P., are the Court's conclusions of law based on the foregoing findings of fact. In short, judgment must enter in favor of the PTO for three reasons: (1) the '809 Application is fatally indefinite; (2) the '809 Application is obvious in light of the prior art; and (3) the '809 Application fails under the doctrine of obviousness-type double patenting.

A. Indefiniteness

First, the PTO argues that a patent should not issue for the '809 Application because the application fails to define AIBW or IBW in such a way that a POSA would be informed, with reasonable certainty, as to the scope of the invention. The PTO raised this argument for the first time in this § 145 litigation; neither the PTAB nor the PTO patent examiner relied on indefiniteness as a basis for rejecting the '809 Application. Nonetheless, the Federal Circuit has made clear that "§ 145 actions open the door to new evidence and arguments," and that the PTO is therefore entitled "to assert its meritorious defenses

under the circumstances," even if the PTO did not present them prior to the § 145 litigation. Hyatt v. Hirshfeld, 998 F.3d 1347, 1363 (Fed. Cir. 2021). Thus, it is appropriate to consider the PTO's indefiniteness argument here, even though the issue was not raised during the patent prosecution process or decided by the PTAB. Because the PTO has raised the issue of indefiniteness for the first time during this litigation, however, the PTO has the burden of proof as to this issue, and must establish indefiniteness by a preponderance of the evidence. See Iancu, 332 F. Supp. 3d at 120 (explaining that the PTO may raise new affirmative defenses in a § 145 proceeding and that the PTO bears the burden of proof on such affirmative defenses); Dome Patent L.P., 799 F.3d at 1377-79 (noting that a preponderance of the evidence standard applies in § 145 proceedings).

The PTO argues persuasively that the '809 Application is fatally indefinite. The indefiniteness analysis begins with the statutory language of § 112 of the Patent Act, which requires that a patent specification "conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor ... regards as the invention." 35 U.S.C. § 112(b) (emphasis added). The § 112 requirement is often "given the shorthand name 'indefiniteness.'" In re Conley, 490 F.2d 972, 975 (C.C.P.A. 1974). And the Supreme Court has explained that the § 112 requirement exists because patents must "be precise enough to afford clear notice of what is claimed, thereby 'appris[ing] the public of what is still open to them.'" Nautilus, 572 U.S. at 899, 134 S.Ct. 2120 (quoting Markman v. Westview Instruments, Inc., 517 U.S. 370, 373, 116 S.Ct. 1384, 134 L.Ed.2d 577 (1996)). As the Supreme Court has further explained, § 112 requires a "reasonable certainty" standard for indefiniteness: a patent is invalid "if its claims, read in light of the specification delineating the patent, and the prosecution history, fail to inform, with reasonable certainty, those skilled in the art about the scope of the invention." Id. at 901, 134 S.Ct. 2120. The Federal Circuit, in applying the Nautilus standard, instructs district courts first to evaluate the language of the claims to determine whether a disputed term is reasonably clear. Berkheimer, 881 F.3d at 1363. Next, the Federal Circuit instructs district courts to look at pertinent intrinsic evidence beyond the language of the claims, including the application's written description and prosecution history. Id. at 1363-64; see also HZNP Meds. LLC v. Actavis Labs. UT, Inc., 940 F.3d 680, 690-91 (Fed. Cir. 2019). Finally, the Federal Circuit instructs district courts to evaluate extrinsic evidence of indefiniteness after evaluating the intrinsic evidence. HZNP Meds., 940 F.3d at 691.

These governing principles make clear that the language of the claims in the '809 Application do not define with reasonable certainty the scope of the claimed invention. ImmunoGen does not dispute that the drug IMGN853 was taught in the prior art, and instead argues that the AIBW dosing regimen in the '809 Application distinguishes the claimed invention. Indeed, all of ImmunoGen's claims in the '809 Application require administering IMGN853 to patients at a dose of 6 mg/kg based on AIBW. But not one of the claims in the '809 Application define AIBW or inform how to calculate a patient's AIBW. Thus, based on the claims of the '809 Application, a POSA would not be reasonably certain of the metes and bounds of the claimed AIBW dosing regimen. In fact, Dr. Figg testified that in reading Claim 1 of the '809 Application, he remained "confused over how to calculate"

AIBW because the claims do not provide a definition.⁶⁸

68 Tr. Nov. 3, 2022 (24:25) (Figg).

The Definitions section of the '809 Application also fails to inform skilled artisans of the scope of the claimed AIBW dosing regimen. Specifically, the Definitions section states that AIBW can be calculated "for example, using the formula ADJ = IBW = 0.4(weight in kg - IBW)." Supra ¶ 43 (emphasis added). In using the phrase "for example," the '809 Application makes clear that there are multiple ways to calculate AIBW, thereby leaving a POSA to guess whether the formula provided is the only one covered by the '809 Application. Indeed, Dr. Figg explained that the Application's inclusion of the term "for example" in the "Definitions" section leaves the definition of AIBW "vague."⁶⁹ And the Federal Circuit has made clear that "open-ended" phrases such as "for example" are "not definitional." Personalized Media Commc'ns, LLC v. Apple Inc., 952 F.3d 1336, 1343 (Fed. Cir. 2020). The Federal Circuit has also noted that "a person of ordinary skill in the art would not understand the 'e.g.' phrase to constitute an exclusive definition." Interval Licensing LLC v. AOL, Inc., 766 F.3d 1364, 1374 (Fed. Cir. 2014). And, of course, "e.g.," means "for example."

69 Tr. Nov. 3, 2022 (25:19-20) (Figg).

Nor does the definition of IBW in the '809 Application, which is used in the AIBW calculation, eliminate the uncertainty about the scope of the claimed dosing method; indeed, it adds even more uncertainty. The Definitions section defines IBW as "an estimate of weight corrected for sex and height, and optionally frame size," and states that IBW "can be calculated, for example, using the formulas IBW = 0.9H-88 (for males) and IBW = 0.9H-92 (for females), wherein H = height in cm." Supra ¶ 44 (emphasis added). Thus, the definition of IBW makes clear that the formulas listed for IBW are but examples of possible methods of calculating IBW, and that a calculation of IBW may optionally account for frame size in an unspecified manner. This only further broadens the realm of possible formulas for IBW. For this reason, Dr. Figg testified that a POSA reading the '809 Application would not know how to calculate IBW because the application's "Definitions" section expressly states that IBW can include an "optional[]" frame size correction and twice uses the term "for example."⁷⁰

70 Tr. Nov. 3, 2022 (25:21-25; 26:1-3) (Figg).

Further compounding the indefiniteness of IBW and AIBW in the '809 Application is the statement in the '809 Application that "IBW ... [is] discussed in more detail in Green and Duffull ... which is herein incorporated by reference in its entirety." Supra ¶ 45. The Green reference provides multiple ways to calculate IBW and multiple correction factors that may be used for AIBW. Thus, by incorporating Green in its entirety, the '809 Application specifically incorporates multiple possible formulas for AIBW and IBW. And as Dr. Figg explained, this prevents a POSA from discerning the scope of the '809 Application with the requisite reasonable certainty, because it leaves skilled artisans "unclear as to which equation to use."⁷¹ See IQASR LLC v. Wendt Corp., 825 F. App'x 900, 905 (Fed. Cir. 2020) (rejecting a claim as indefinite where it included "multiple layers of definitions ... all open-ended and non-limiting"); Teva Pharms. USA, Inc. v. Sandoz, Inc., 789 F.3d 1335, 1345 (Fed. Cir. 2015) (holding that claims were indefinite where the claims required a specific measurement and more than one measurement method could be used, but the patent

71 Tr. Nov. 3, 2022 (26:11-15) (Figg).

provided no guidance on which specific method to use).

In addition to this intrinsic evidence, the PTO also introduced extrinsic evidence further indicating that the '809 Application fails to define AIBW and IBW with reasonable certainty. As of October 8, 2013, the prior art disclosed at least eleven formulas for calculating IBW. See supra ¶ 58. And the art also disclosed a wide range of correction factors used to calculate AIBW for different drugs: although 0.4 is often used as a standard correction factor, the correction factor values can range from 0.2 to 0.9. See supra ¶ 56. Even ImmunoGen's expert, Dr. Shah, admitted that the choice in correction factor can lead to a "huge" difference in the resulting dose.⁷² Thus, as Dr. Figg explained, the numerous different correction factors reported in the literature would leave a POSA uncertain as to which correction factor to use when calculating AIBW for IMGN853, thereby leaving indefinite the scope of the claimed invention in the '809 Application. See supra ¶ 59. Further adding to this confusion and uncertain scope of the claimed invention, the record reflects that some drugs have more than one reported correction factor in the literature. For example, as Dr. Figg pointed out, there are at least three correction factors for the drug tobramycin. See supra ¶ 57. Thus, because of the wide range of correction factors used in drug dosing for AIBW and the fact that some drugs have more than one reported correction factor for dosing, a POSA would not know whether the 0.4 correction factor used in the '809 Application is the only correction factor for IMGN853, or whether there are other potential correction factors claimed by the '809 Application. This extrinsic evidence therefore leaves a POSA to wonder just how many correction factors and AIBW formulas are encompassed by the '809 Application, and thus this evidence further indicates that the '809 Application must be rejected as impermissibly indefinite.

72 Tr. Nov. 2, 2022 (143:8-12) (Shah).

Seeking to avoid this conclusion, ImmunoGen argues that a POSA would understand that the definition of AIBW is limited to the specific formula provided in Example 4 of the '809 Application. ImmunoGen argues that the formulas for AIBW and IBW set forth in Example 4 are the same AIBW and IBW formulas used "for example" in the Definitions section, used by researchers during the clinical trials of IMGN853, and referenced by the PTO during patent prosecution. Thus, ImmunoGen contends that a POSA would not be confused by the use of "for example" in the Definitions section, and would know to apply the formulas in Example 4 rather than other formulas from the literature that were never associated with IMGN853.

ImmunoGen's argument that Example 4 clarifies the definition of AIBW and IBW fails to persuade. To begin with, the '809 Application specifically denies that its examples limit the scope of its claims; rather, the application states that "[i]t is understood that the examples and embodiments described herein are for illustrative purposes only and that that various modifications or changes in light thereof will be suggested to persons skilled in the art." Supra ¶ 47. Furthermore, the conclusion of the "Examples" section of the '809 Application also describes the examples as non-limiting, stating that "the breadth and scope of the present invention should not be limited by any of the above-described exemplary embodiments, but should be defined only in accordance with the following claims and their equivalents." Supra ¶ 48. Based on this disclaimer language and the numerous equations and correction factors disclosed in the literature, Dr. Figg explained

that a POSA would not understand that the formula for AIBW listed in Example 4 was the only formula for AIBW claimed in the '809 Application.⁷³

73 Tr. Nov. 3, 2022 (27:9-22) (Figg).

Federal Circuit precedent also makes clear that the otherwise indefinite claims of the '809 Application should not be limited by the formulas provided in Example 4. Specifically, the Federal Circuit has "repeatedly warned against confining the claims" of an application to the "specific embodiments" described in the specification. Phillips v. AWH Corp., 415 F.3d 1303, 1323 (Fed. Cir. 2005) (en banc). And "even if a specification has only one embodiment, its claims will not be confined to that example unless the patentee has demonstrated a clear intention to limit the claim scope using words or expression of manifest exclusion or restriction." Aria Diagnostics, Inc. v. Sequenom, Inc., 726 F.3d 1296, 1301 (Fed. Cir. 2013) (quoting Liebel-Flarsheim Co. v. Medrad, Inc., 358 F.3d 898, 906 (Fed. Cir. 2004)). Not so here: the '809 Application contains no words expressly limiting its scope to Example 4, and indeed contains multiple disclaimers that the claims of the application should not be limited by the embodiments. ImmunoGen cannot have it both ways. As Aria Diagnostics teaches, a patentee in suing for infringement is not limited to the patent's embodiments; therefore, the same patentee cannot rely on the embodiments to rescue the patent or patent application from fatal indefiniteness. Thus, both Federal Circuit precedent and the '809 Application itself preclude ImmunoGen from relying on Example 4 to limit the scope of the claimed dosing invention.

ImmunoGen also argues that the prosecution history of the '809 Application further indicates that a POSA would understand which AIBW formula to use for IMGN853. In support of this argument, ImmunoGen points to the fact that neither the patent examiner nor the PTAB ever raised an indefiniteness rejection for the '809 Application. This argument is plainly foreclosed by Federal Circuit caselaw making clear that the PTO is entitled to raise new grounds for rejecting a patent application during § 145 proceedings. See Hyatt, 998 F.3d at 1363 ("§ 145 actions open the door to new evidence and arguments."). The fact that the PTO missed the indefiniteness issue during examination does not foreclose the issue now. Otherwise, the rule allowing the PTO to raise a new rejection in § 145 actions would be a dead letter.

ImmunoGen also contends that the prosecution history indicates that the '809 Application is not indefinite because during examination, the patent examiner, the PTAB, ImmunoGen, and ImmunoGen's expert all understood that AIBW meant the precise formula provided in Example 4. This argument overstates the record and fails to save the '809 Application. To begin with, the patent examiner merely applied the Example 4 AIBW formula in proposing a hypothetical. See Administrative Record ("AR") at 6102. This only demonstrates that the examiner understood that AIBW included the formula recited in Example 4, but by no means does this indicate that the examiner understood the Example 4 formula to be the only definition of AIBW claimed in the '809 Application.

Moreover, although ImmunoGen contends that the PTAB applied the Example 4 formula, in fact, the PTAB merely quoted the Definitions section of the '809 Application—the section which includes two uses of the term "for example" and one other use of "optionally." See supra ¶ 52; AR 6180. To be sure, ImmunoGen and ImmunoGen's expert during the prosecution

process referenced Example 4's formula for AIBW during the patent examination process, see AR 4564-92, 5703-07, but ImmunoGen's own statements during prosecution are far from dispositive on the issue of indefiniteness. Although the Federal Circuit has held that statements during prosecution by the applicant can inform the meaning of the claims and are relevant to the indefiniteness inquiry, see Am. Piledriving Equip., Inc. v. Geoquip, Inc., 637 F.3d 1324, 1336 (Fed. Cir. 2011); In re Packard, 751 F.3d 1307, 1312 (Fed. Cir. 2014), these cases do not hold that such evidence is determinative on the issue. Instead, the indefiniteness analysis requires viewing the patent's claims "in light of the specification and prosecution history" as a whole, Nautilus, 572 U.S. at 910, 134 S.Ct. 2120, which here indicates that the '809 Application encompasses many different formulas for AIBW and IBW.

In conclusion, the '809 Application is fatally indefinite because it fails to define AIBW or IBW anywhere in its claims. Furthermore, although the Definitions section lists a formula for AIBW and IBW, the formulas are preceded by the phrase "for example,"⁷⁴ and the Definitions section incorporates the Green reference, which lists various IBW formulas and AIBW correction factors. Finally, although Example 4 lists specific formulas for AIBW and IBW, the '809 Application expressly disclaims that its examples do not limit the scope of the claims. The '809 Application therefore encompasses multiple different AIBW and IBW formulas and fails to define, with the requisite certainty, the scope of the invention. Because ImmunoGen relies on the AIBW limitation to distinguish the '809 Application from the prior art, it is especially critical that the application make clear the metes and bounds of the term. As demonstrated at trial, the '809 Application plainly fails to do so, leaving skilled artisans to guess about the invention's scope. For that reason, the '809 Application must be rejected as indefinite.

74 The plain English meaning of the phrase "for example" means that the phrase provides just one example out of many possible options. A POSA is not at liberty to ignore the plain meaning of the phrase. See Interval Licensing LLC, 766 F.3d at 1374 ("Had the phrase been cast as a definition instead of as an example ... then it would help provide the clarity that the specification lacks. But as the specification is written ... a person of ordinary skill in the art would not understand the "e.g." phrase to constitute an exclusive definition[.]").

B. Obviousness

Next, the parties dispute whether the patent examiner and the PTAB properly determined that the claims of the '809 Application are obvious in light of the prior art. The PTO argues that the prior art taught using IMGN853 to treat ovarian and peritoneal cancers and taught dosing based on AIBW for a variety of drugs. According to the PTO, a POSA would have been motivated to combine the known IMGN853 drug with the known AIBW dosing methodology and would have had a reasonable expectation of success in doing so. In response, ImmunoGen argues that (i) the '809 Application is not obvious because there are a number of differences between the prior art and the claimed invention, (ii) there was no motivation to combine the known AIBW dosing methods with IMGN853, and (iii) a POSA would not have a reasonable expectation of success in using AIBW to dose IMGN853. Finally, the parties dispute whether secondary considerations such as unexpected results, industry praise, and copying indicate that the '809 Application is obvious.

The Patent Act requires that patents address "non-obvious subject matter." 35 U.S.C. § 103. As the Supreme Court

explained in Graham v. John Deere Co., 383 U.S. 1, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966), obviousness is a question of law based on underlying facts. Id. at 17-18, 86 S.Ct. 684. Specifically, Graham teaches that the obvious analysis consists of three factual inquiries:

(1) determining the scope and content of the prior art;

(2) ascertaining the differences between the prior art and the claims at issue; and

(3) identifying the level of ordinary skill in the art at the time the invention was made.

Id. at 17, 86 S.Ct. 684. If the "difference between the subject matter sought to be patented and the prior art" is "such that the subject matter as a whole would have been obvious at the time to a person skilled in the art, then the subject matter cannot be patented." Id. at 15, 86 S.Ct. 684. The Graham Court also explained that other objective "secondary considerations" or "indicia of obviousness" should be considered in the obviousness analysis, as these indicia can "give light to the circumstances surrounding the origin of the subject matter sought to be patented." Id. at 17-18, 86 S.Ct. 684. Such secondary considerations include "commercial success, long felt but unsolved needs, [and] failure of others." Id. at 17, 86 S.Ct. 684.

More recently, the Supreme Court clarified in KSR International Co. v. Teleflex, Inc., 550 U.S. 398, 127 S.Ct. 1727, 167 L.Ed.2d 705 (2007), that a patent for an invention "composed of several elements is not proved obvious merely by demonstrating that each of its elements was, independently, known in the prior art." Id. at 418, 127 S.Ct. 1727. Instead, the reviewing court must "determine whether there was an apparent reason to combine the known elements in the fashion claimed by the patent at issue." Id. Thus, after KSR, a patent claiming a combination of previously disclosed elements is not invalid for obviousness unless the reviewing court determines that a POSA would have been motivated to combine the prior art references to create the claimed invention. In other words, even though "common sense directs one to look with care at a patent application that claims as innovation the combination of two known devices," courts must take care to "identify a reason that would have prompted a person of ordinary skill ... to combine the elements in the way the claimed new invention does," because "inventions in most, if not all, instances rely upon building blocks long since uncovered." Id. at 418, 127 S.Ct. 1727. In applying KSR, the Federal Circuit has explained that the factfinder must consider whether (i) "a skilled artisan would have been motivated to combine the teachings of the prior art references to achieve the claimed invention," and (ii) whether "the skilled artisan would have had a reasonable expectation of success in doing so." Novartis Pharm. Corp. v. West-Ward Pharm. Int'l Ltd., 923 F.3d 1051, 1059 (Fed. Cir. 2019). Both the "presence or absence of a motivation to combine references" and the "presence or absence of a reasonable expectation of success" are "pure question[s] of fact." Id.

As an initial matter, it is appropriate to note that a district court can determine whether the claims within a patent application are obvious even after determining that the same claims are indefinite. The Federal Circuit has recognized that "[t]he questions [of obviousness and indefiniteness] are different" and the finding that a claim is indefinite does not "imply that it is incapable of being compared to prior art to determine if one of its alternatives is anticipated or would have been obvious on the grounds asserted." Cochlear Bone Anchored Sols. AB v. Oticon Med. AB, 958 F.3d 1348, 1360 (Fed.

Cir. 2020). Thus, the indefiniteness of the '809 Application at issue here does not preclude a comparison of the '809 Application to the prior art. And for the purpose of the obviousness analysis, ImmunoGen may be taken at its word that the AIBW formula ImmunoGen intended to patent is the one in Example 4 of the '809 Application. See id. (holding that a court may focus on one of an application's alternatives in determining obviousness).

Each of the elements of the Graham and KSR obviousness inquiry are analyzed in turn below. In short, the patent examiner and the PTAB correctly concluded that the '809 Application is obvious. At the bench trial, the PTO persuasively demonstrated that the elements of the '809 Application were disclosed in the art prior to October 8, 2013, that a POSA would have been motivated to combine those references and dose IMGN853 at 6 mg/kg AIBW, and that in doing so, a skilled artisan would have a reasonable expectation of success. Finally, secondary considerations of obviousness weigh in the PTO's favor. Thus, ImmunoGen has failed to show that the '809 Application is not obvious in violation of § 103; ImmunoGen is therefore not entitled to a patent for the invention claimed in the '809 Application. See Disney Enterprises, 940 F. Supp. 2d at 292 ("[The applicant] bears the burden of proving by a preponderance of the evidence that it is entitled to a ... patent covering the pending claims in the ... application.").

1. Level of Ordinary Skill in the Art

The first step in the obviousness analysis is to identify the level of ordinary skill in the art at the time the invention was made. See Graham, 383 U.S. at 17, 86 S.Ct. 684. ImmunoGen and the PTO used slightly different formulations to define the qualifications of a POSA at the time the '809 Application was filed in October 2013. Despite this, all three experts agreed that their definitions are "very similar," and the three experts also stated that their opinions would not change if the other party's definition were adopted.⁷⁵ It is therefore appropriate to adopt Dr. Figg's definition of a POSA. Thus, at the time of the '809 Application, a POSA would have an advanced degree in clinical pharmacy or clinical pharmacology and at least two years of experience in the field of clinical pharmacy, clinical pharmacology, or drug development with an emphasis on oncology agents, or would be a clinical oncologist with experience in drug development and similar years of experience. See supra ¶ 39.

75 See Tr. Nov. 2, 2022 (114:20-23) (Dr. Shah); Tr. Nov. 1, 2022 (15912-24) (Dr. Tolcher); Tr. Nov. 3, 2022 (13:19) (Dr. Figg).

2. Scope and Content of the Prior Art

The next step is to determine from the trial evidence the scope and content of the art as of October 8, 2013, the priority date of the '809 Application. See Graham, 383 U.S. at 17, 86 S.Ct. 684. The prior art is set forth in detail in the above findings of fact. See supra ¶¶ 61-65. Simply put, not only was IMGN853 known in the art prior to October 8, 2013, but it was also known that the administration of IMGN853 had the potential to cause ocular toxicity in humans because IMGN853 contains the DM4 payload. AIBW was also known in the prior art, including as a method for dosing multiple antibiotics and at least six anticancer treatments. Furthermore, AIBW had been used to dose both large and small molecule drugs. The specific AIBW formula listed in the "Definitions" section and Example 4 of the '809 Application was also known in the prior art, as Drs. Ponte and Lutz testified that

they found the formula on a nutrition website. Finally, the 0.4 correction factor used in the '809 Application's AIBW formula was not only disclosed in the prior art but also was commonly used as a standard correction factor. Based on these prior art references, a POSA would have been familiar with AIBW as a potential dosing method for immunoconjugates like IMGN853 as of October 8, 2013.

3. Differences Between the Prior Art and the Claims at Issue

Graham instructs that the next step in the obviousness analysis is to determine the differences between the prior art and the claims at issue. See Graham, 383 U.S. at 17, 86 S.Ct. 684. These differences are described in detail in the findings of fact set forth above, supra ¶¶ 66-72.

In arguing that the '809 Application is not obvious, ImmunoGen points to six differences between the prior art and the '809 Application:

(1) AIBW dosing had never been used to dose an immunoconjugate.

(2) The prior art does not specifically disclose dosing IMGN853 at 6 mg/kg AIBW.

(3) The prior art focuses on AIBW dosing in obese or overweight patients, but the '809 Application describes using AIBW dosing for all patients regardless of weight.

(4) No prior art reference discloses that IMGN853 caused ocular toxicity in humans.

(5) No prior art reference specifically recommends using AIBW dosing in order to overcome ocular toxicity.

(6) No prior art reference discloses a correlation between the PK parameters for early exposure of IMGN853 and the occurrences of ocular toxicity.

Simply put, none of these purported differences between the prior art and the '809 Application supports that the application is unobvious. First, ImmunoGen points out that AIBW dosing had never been used to dose an immunoconjugate, but the trial evidence persuasively reflects that AIBW dosing had been used in a variety of anticancer drugs before October 2013, including both large and small molecule drugs. See supra ¶¶ 62-63, 68. Thus, as Dr. Figg testified, a POSA would have been familiar with AIBW dosing as a potential dosing method for immunoconjugates, even though the prior art does not disclose specifically that AIBW had previously been used to dose immunoconjugates. See supra ¶¶ 63, 76; see also Tr. Nov. 3, 2022 (41:19-20) (Figg) ("[I]t is common knowledge to [use AIBW] in the field of drug development of anticancer agents.").

Second, although the prior art does not disclose dosing IMGN853 at 6 mg/kg AIBW, the prior art does disclose dosing IMGN853 at around 6 mg/kg TBW. As explained infra in Section III.B.4, a POSA would have been motivated to combine the known 6 mg/kg TBW dose with the known AIBW dosing methodology to arrive at a dose of 6 mg/kg AIBW.

Third, although most of the prior art references focus on AIBW dosing in the context of overweight or obese patients, the Rudek reference discusses using AIBW to dose an anticancer drug for every type of patient, independent of weight. See supra ¶ 69.

Fourth, although no prior art reference specifically discloses that IMGN853 causes ocular toxicity in humans, it was well understood that the administration of immunoconjugates containing the DM4 payload could result in ocular toxicity in human patients. See supra ¶¶ 8, 74.

Fifth, although no prior art reference explicitly recommends AIBW dosing to overcome ocular toxicity, the Alffenaar reference reports testing AIBW dosing as a possible solution for ocular toxicity associated with the antibiotic ethambutol. See

supra ¶ 62. Thus, even though Alffenaar ultimately recommended IBW dosing for ethambutol, the reference shows that AIBW dosing was disclosed in the prior art as a possible solution for ocular toxicity.

Sixth, despite the fact that the prior art does not disclose ImmunoGen's PK data for IMGN853 indicating a correlation between early exposure and ocular toxicity, Dr. Figg testified that in 2013, such PK testing was routinely conducted in the course of drug development.⁷⁶ Thus, any POSA experimenting with IMGN853 would have conducted similar PK analyses.

76 Tr. Nov. 3, 2022 (57:4-6) (Figg).

In summary, none of ImmunoGen's asserted differences between the prior art and the invention claimed in the '809 Application is material or indicates that the claims at issue are unobvious. And as discussed below, a POSA would have been motivated to combine the prior art references, which disclose using AIBW dosing for various drugs, to arrive at the claimed dose in the '809 Application of 6 mg/kg AIBW.

4. Motivation to Combine the Prior Art Elements

The Federal Circuit has made clear that an "invention is not obvious simply because all of the claimed limitations were known in the prior art at the time of the invention." Forest Labs., LLC v. Sigmapharm Labs., LLC, 918 F.3d 928, 934 (Fed. Cir 2019). Instead, the Federal Circuit directs district courts to ask "whether there is a reason, suggestion, or motivation in the prior art that would lead one of ordinary skill in the art to combine the references." Id. (quoting Smiths Indus. Med. Sys., Inc. v. Vital Signs, Inc., 183 F.3d 1347, 1356 (Fed. Cir. 1999)). Thus, the next issue is whether a POSA would have had a motivation to combine the prior art references in order to arrive at the AIBW dosing regimen disclosed in the '809 Application. This motivation "can be found explicitly or implicitly in the prior art references themselves, in market forces, in design incentives, or in 'any need or problem known in the field of endeavor at the time of invention and addressed by the patent.'" Id. (quoting Arctic Cat Inc. v. Bombardier Recreational Prods. Inc., 876 F.3d 1350, 1359 (Fed. Cir. 2017)).

Here, the trial record reflects that a POSA as of the priority date of the '809 Application would have been motivated to combine the prior art references. To begin with, a POSA would understand the potential risk of ocular toxicity given that IMGN853 contains the DM4 payload, which for over twenty years was known to cause ocular toxicity in humans. See supra ¶ 8. Given this knowledge, a skilled artisan would monitor for ocular toxicity when conducting clinical trials of IMGN853 because it contains the DM4 payload. Indeed, Dr. Lutz testified that he monitored for ocular toxicity during the clinical trials of IMGN853 based on the known history of the DM4 payload.⁷⁷ Once a POSA observed ocular toxicity from IMGN853 in the ordinary course of human trials, the skilled artisan would then experiment with changing the dose to reduce toxicity. See supra ¶ 75. Even ImmunoGen's expert, Dr. Tolcher, testified that although the exact mechanism by which the DM4 payload causes ocular toxicity is unknown, the toxicity "is usually dose-related."⁷⁸ Thus, as Dr. Figg explained, when confronted with ocular toxicity, a skilled artisan would look to changing the dose of IMGN853.⁷⁹

77 Tr. Nov. 1, 2022 (110:16-22) (Lutz).

78 Tr. Nov. 1, 2022 (164:23-25; 165:5-8) (Tolcher).

79 Tr. Nov. 3, 2022 (56:19-23) (Figg).

Once a POSA decided to experiment with the dosing methodology for

IMGN853, the skilled artisan would try AIBW dosing, given that AIBW dosing was well known in the prior art. See supra ¶ 76. And the skilled artisan would look to prior art references pertaining to both large and small molecule drugs, even though immunoconjugates are large molecule drugs. See supra ¶ 77. Thus, the prior art references discussing the use of AIBW in small molecule drugs would motivate a POSA to experiment with AIBW for IMGN853, as would the two references teaching the use of AIBW for the transplantation of CD34+ cells—which are thousands of times larger than immunoconjugates. See supra ¶¶ 68, 77. Moreover, Alffenaar and van der Werk's reference would have provided additional motivation to a POSA to try AIBW dosing in order to reduce the ocular toxicity associated with IMGN853. The Alffenaar reference describes that before the priority date of the '809 Application, researchers considered using AIBW to reduce ocular toxicity in patients receiving ethambutol. See supra ¶¶ 62, 78. This reference would suggest to a skilled artisan that AIBW dosing may be used as a potential solution for ocular toxicity and would provide motivation and reason for the skilled artisan to experiment with AIBW in attempting to eliminate or ameliorate the toxicities associated with IMGN853.

Once a POSA decided to experiment with AIBW dosing based on the prior art references, the skilled artisan would start with doses of around 5 mg/kg or 6 mg/kg AIBW and then determine the precise dose based on routine optimization. Lutz '282 focuses on dosing IMGN853 at around 6 mg/kg TBW, and the 2013 ASCO abstract disclosed that IMGN853 had been tested on humans at a dose of 5 mg/kg TBW. See supra ¶¶ 61, 81. Thus, a skilled artisan seeking to determine the proper AIBW dose would start with doses of either 5 mg/kg or 6 mg/kg AIBW and determine the best dose through experimentation, which is precisely what ImmunoGen's researchers did. See supra ¶ 26. Furthermore, a POSA would be motivated to use the 0.4 correction factor, because as Dr. Figg explained, 0.4 was used as a standard correction factor and hence researchers often start with 0.4 when experimenting with AIBW dosing. See supra ¶ 82.

Finally, the fact that other researchers were contemporaneously motivated to try AIBW dosing in order to reduce toxic side effects provides further support for the conclusion that a POSA as of October 8, 2013 would have been motivated to experiment with AIBW dosing for IMGN853. The Rosenblat reference demonstrates that around the same time that Drs. Lutz and Ponte were developing the '809 Application, other researchers were testing AIBW dosing to ameliorate toxic side effects associated with a radioimmunoconjugate. See supra ¶ 79. Although the Rosenblat reference is not in the prior art because it was not published until 2022, it discloses an "independently made, simultaneous invention[]" made "within a comparatively short space of time," which the Federal Circuit has explained can provide "persuasive evidence that the claimed [invention] was the product only of ordinary mechanical or engineering skill." Geo. M. Martin Co., 618 F.3d at 1305 (quoting Concrete Appliances Co. v. Gomery, 269 U.S. 177, 184, 46 S.Ct. 42, 70 L.Ed. 222 (1925)). The Rosenblat reference therefore provides "persuasive evidence" that a POSA would have been motivated to combine the prior art references and experiment with AIBW dosing. Id.

Seeking to avoid this conclusion, ImmunoGen argues that a POSA would not have been motivated to experiment with AIBW dosing for IMGN853 in order to eliminate ocular toxicity because Lutz '282 describes that IMGN853 did not result in ocular toxicity when administered to rabbits. According

to ImmunoGen, based on the rabbit studies described in Lutz '282, a skilled artisan would believe that the use of the charged linker in IMGN853 solved the problem of ocular toxicity and would therefore have no reason to try AIBW dosing. In support, ImmunoGen cites Federal Circuit precedent holding that "where a problem was not known in the art, the solution to that problem may not be obvious, because 'ordinary artisans would not have thought to try at all because they would not have recognized the problem.'" Forest Labs., 918 F.3d at 935 (quoting Leo Pharm. Prods., 726 F.3d at 1357).

This argument misinterprets Forest Laboratories and ignores pertinent facts in this case. In Forest Laboratories, the Federal Circuit explained that a drug patent disclosing a dosing method was not obvious where the inventors discovered an entirely unknown side effect—cardiotoxic events—and developed the claimed dosing method as a solution to the problem. Id. In that case, the fact that the drug at issue caused cardiotoxic events was never disclosed in the prior art. For that reason, Forest Laboratories is inapposite here, where, as ImmunoGen admits,⁸⁰ it was well known for over twenty years the administration of immunoconjugates like IMGN853 containing the DM4 payload caused ocular toxicity. Unlike the researchers in Forest Laboratories, ImmunoGen does not get credit for "discovering" the ocular toxicity associated with IMGN853, because the discovery itself would have been obvious to a POSA given the well-known history of the DM4 payload.

80 See Joint Submission of Stipulated Facts, Dkt. 127 at ¶ 29.

Contrary to ImmunoGen's argument, the rabbit studies described in Lutz '282 do not alter the result that a POSA as of the priority date of the '809 Application would have known about the potential for ocular toxicity from IMGN853. Both Dr. Lutz and Dr. Figg testified that despite the suggestion in Lutz '282 that using a charged linker instead of an uncharged one in IMGN853 eliminated ocular toxicity when IMGN853 was tested in rabbits, it is well known that side effects may occur in humans which do not appear during animal testing. See supra ¶ 14. Researchers entering clinical trials of IMGN853 would therefore understand the need to monitor for ocular toxicity in human patients regardless of the results of the preclinical rabbit studies described in Lutz '282. That others did not discover the fact that IMGN853 caused ocular toxicity is presumably because ImmunoGen was the only one who stood to gain from experimenting with its already-patented IMGN853 drug. See Acorda Therapeutics, Inc. v. Roxane Labs., Inc., 903 F.3d 1310, 1339 (Fed. Cir. 2018) ("[A] blocking patent diminishes possible rewards from a non-owner's ... investment activity aimed at an invention whose commercial exploitation would be infringing[.]"); see also supra note 67. ImmunoGen's argument that a skilled artisan would not have been motivated to experiment with AIBW dosing based on Lutz '282 therefore fails.

ImmunoGen next argues that even if a POSA would have recognized a need to modify the dosing in Lutz '282, nothing in the prior art would motivate a skilled artisan to try the '809 Application dose of 6 mg/kg AIBW using the specific AIBW formula recited in Example 4. This argument fails to persuade for two reasons. First, a POSA would have been motivated to try the recited dose of 6 mg/kg AIBW based on two prior art references: Lutz '282 and the 2013 ASCO abstract.⁸¹

81 In ImmunoGen's Response to the PTO's Proposed Findings of Fact and Conclusions of Law, ImmunoGen argues that the PTO waived its argument that the '809 Application is obvious in light of the 2013 ASCO abstract because the PTO did not raise the argument in its pre-trial briefs, discovery responses, or expert testimony. In support, ImmunoGen cites only a non-binding decision from the District of Delaware. See ImmunoGen's Response to Defendant's Proposed Findings of Fact and Conclusions of Law, Dkt. 144 at 3 (citing Allergan, Inc. v. Barr Labs., Inc., 808 F. Supp. 2d 715, 735 (D. Del. 2011)). ImmunoGen's waiver argument fails; the PTO did not waive its argument about the 2013 ASCO abstract. The 2013 ASCO abstract was admitted as an exhibit at trial without any objection from ImmunoGen, and Dr. Figg testified about the abstract at trial, explaining that the abstract discloses that IMGN853 was tested in humans at a dose of 5 mg/kg TBW. See Tr. Nov. 3, 2022 (35:1-8; 65:5-9) (Figg). And ImmunoGen even stipulated that the 2013 ASCO abstract is prior art in this case, further undermining ImmunoGen's assertion that the PTO waived arguments about the 2013 ASCO abstract. See Joint Submission of Stipulated Facts, Dkt. 127 at ¶¶ 56-57.

As Dr. Figg explained, a POSA would be motivated to try a dose of 5 mg/kg AIBW because the 2013 ASCO abstract discloses that as of the priority date of the '809 Application, patients had received IMGN853 at a dose of 5 mg/kg TBW.⁸² Based on the 2013 ASCO abstract, a skilled artisan experimenting with AIBW would start by dosing around 5 mg/kg AIBW. Similarly, as Dr. Figg explained, Lutz '282 focused on doses around 6 mg/kg TBW.⁸³ Thus, based on the prior art, a POSA experimenting with AIBW dosing for IMGN853 would start by dosing around 5 to 6 mg/kg AIBW and would then escalate the dose in order to find the maximum tolerated dose that would minimize ocular toxicity. Indeed, that is precisely what ImmunoGen did: 5 mg/kg and 6 mg/kg are the only two AIBW doses that ImmunoGen ever tested. See supra ¶ 26. Moreover, Ab '528 and Carrigan '675, both of which involve IMGN853, teach that an "administering physician can easily determine optimum dosages, dosing methodologies and repetition rates." Supra ¶ 61. Thus, the prior art teaches that a POSA would start by dosing IMGN853 at 5 to 6 mg/kg AIBW and, like ImmunoGen, determine the optimum dose based on routine experimentation with dosing.⁸⁴

82 Tr. Nov. 3, 2022 (65:5-9) (Figg).

83 Tr. Nov. 3, 2022 (33:7-16) (Figg).

84 This conclusion is supported by the Federal Circuit's decision in Boehringer Ingelheim Pharms. Inc. v. Mylan Pharms., Inc., 803 F. App'x 397 (Fed. Cir. 2020). There, the Federal Circuit affirmed a finding of obviousness of a dosing methodology where the district court determined that a POSA would have a reasonable expectation of arriving at the claimed doses "through routine experimentation." Id. at 402.

Additionally, a skilled artisan would have been motivated to try the specific AIBW formula described in Example 4—which Drs. Lutz and Ponte did not invent, but found on a nutrition website—because the Example 4 formula follows the standard template for AIBW formulas and utilizes the most common correction factor. As Dr. Figg explained, AIBW formulas typically follow the following template, where "CF" stands for "correction factor": AIBW = IBW + CF (weight in kg - IBW). See supra ¶ 55.⁸⁵ The Example 4 formula follows this template, but uses 0.4 as the correction factor. As Dr. Figg testified, 0.4 has been known as a standard correction factor since the 1980s. See supra ¶ 82. Indeed, as Dr. Figg also explained, researchers often start with 0.4 as a correction factor when experimenting with AIBW dosing. Id. Thus, based on the information in the prior art, a POSA would have been motivated to try the recited dose of 6 mg/kg AIBW using the formula disclosed in Example 4 of the '809 Application. ImmunoGen also argues that two references would dissuade a skilled artisan from experimenting with AIBW dosing for IMGN853: (i) the 2012 ASCO panel, which recommended using TBW instead of AIBW for chemotherapy agents; and (ii) the Green reference, which discusses AIBW but ultimately recommends dosing based on lean body weight. In support, ImmunoGen cites Federal Circuit caselaw explaining that the district court must consider prior art references "that would lead away from the invention in suit." Panduit Corp. v. Dennison Mfg. Co., 810 F.2d 1561, 1568 (Fed. Cir. 1987). But the Federal Circuit has also made clear that pointing to "isolated prior art disclosures" that may lead away from the claimed invention is not sufficient: instead, the district court must determine obviousness "[v]iewing the prior art as a whole." Merck & Cie v. Gnosis S.P.A., 808 F.3d 829, 834 (Fed. Cir. 2015). The 2012 ASCO panel and Green are only "isolated prior art disclosures," id., and, in any event, would not discourage a skilled artisan from experimenting with AIBW dosing for IMGN853. Although Green ultimately recommends lean body weight, the reference recognizes that many drugs are in fact dosed based on AIBW. And as Dr. Figg explained, the 2012 ASCO panel would not have dissuaded an oncologist from using AIBW dosing when treating patients.⁸⁶ Even ImmunoGen's expert, Dr. Tolcher, testified that he "can't disagree" with the statement that the ASCO recommendation not to use AIBW dosing "isn't very persuasive."⁸⁷ Additionally, the 2012 ASCO panel only recommended TBW for chemotherapy drugs, and IMGN853 is not a chemotherapy drug. Finally, even after the 2012 ASCO panel, the FDA did not change its requirement of using AIBW dosing for obese patients for the chemotherapy drug Busulfex. See supra ¶ 80. This demonstrates that even the FDA was not persuaded by the ASCO recommendation not to use AIBW dosing. Thus, ample evidence indicates that neither Green nor the 2012 ASCO recommendation would prevent a skilled artisan from experimenting with AIBW dosing for IMGN853.

85 See also Tr. Nov. 3, 2022 (14:17-21) (Figg).

86 Tr. Nov. 3, 2022 (54:14-19) (Figg).

87 Tr. Nov. 1, 2022 (189:14-17) (Tolcher).

Finally, ImmunoGen argues that even if using AIBW dosing for obese patients is obvious in light of the prior art, the '809 Application is not obvious because it recites using AIBW for all patients regardless of weight, and nothing in the prior art suggests applying AIBW dosing to underweight or cachectic patients. This argument also fails to persuade. To begin with, ImmunoGen misreads the prior art. Although most of the prior art discusses using AIBW in the context of obese patients, the Rudek reference describes using AIBW dosing for an anticancer treatment without reference to patient body weight. See supra ¶ 69. Furthermore, ImmunoGen's argument that the '809 Application applies AIBW to all patients, including underweight patients, does not render the '809 Application unobvious. This is so because for patients who weigh exactly their ideal body weight, a dose of 6 mg/kg AIBW is identical to a dose of 6 mg/kg TBW. See supra ¶ 95. Thus, the '809 Application in effect seeks to cover a dose that was already disclosed in the prior art. And in these circumstances, the Federal Circuit has long held that claims broad enough to cover obvious subject matter are unpatentable even if they read on unobvious claims that would otherwise be patentable. In re Cuozzo Speed Techs., LLC, 793 F.3d 1268, 1281 (Fed. Cir. 2015), aff'd sub nom. Cuozzo Speed Techs., LLC v. Lee, 579 U.S. 261, 136 S.Ct. 2131, 195 L.Ed.2d 423 (2016) (citations omitted) ("It is a long-established rule that claims which are broad

enough to read on obvious subject matter are unpatentable even though they also read on nonobvious subject matter."). This principle applies here, for as Dr. Figg testified and Drs. Shah and Tolcher admitted, when a patient weighs exactly her ideal body weight, she will receive the same amount of IMGN853 regardless of whether she is dosed based on TBW or AIBW.⁸⁸ And because dosing IMGN853 at 6 mg/kg TBW is disclosed in Lutz '282, the claims of the '809 Application, if patented, would prevent doctors from practicing what is already disclosed in the prior art for patients at their ideal body weight. Thus, because for patients at their ideal body weight the IMGN853 dose under the '809 Application is exactly what it would be under the TBW dosing disclosed in Lutz '282, the claims of the '809 Application "are broad enough to read on obvious subject matter" and are therefore invalid as obvious. Cuozzo, 793 F.3d at 1281.

88 Tr. Nov. 2, 2022 (53:22-23) (Tolcher); Tr. Nov. 2, 2022 (171:21-172:7) (Shah); Tr. Nov. 3, 2022 (29:10-16) (Figg).

In conclusion, a POSA as of October 8, 2013 would have been motivated to combine the prior art references to arrive at the '809 Application's recited dose of 6 mg/kg AIBW for IMGN853, and all of ImmunoGen's arguments to the contrary fail to persuade. Furthermore, the '809 Application is obvious and unpatentable because in circumstances where a patient weighs exactly her ideal body weight, the claims of the '809 Application cover a dose already disclosed in the prior art.

5. Reasonable Expectation of Success

To support a finding of obviousness, the evidence must also show that a "skilled artisan would have had a reasonable expectation of success" from "combin[ing] the teaching[s] of the prior art references to achieve the claimed invention." In re Cyclobenzaprine Hydrochloride Extended-Release Capsule Patent Litig., 676 F.3d 1063, 1068-69 (Fed. Cir. 2012). In other words, even if it is "obvious to experiment" by combining elements known in the prior art, a claimed invention is not obvious unless there is something in the prior art "to indicate that a skilled artisan would have had a reasonable expectation that such an experiment would succeed in being therapeutically effective." Id. at 1070. On the one hand, where a "skilled artisan merely pursues 'known options' from a 'finite number of identified, predictable solutions,' the resulting invention is obvious." Id. (quoting KSR, 550 U.S. at 421, 127 S.Ct. 1727). But on the other hand, where the "prior art gave either no indication of which parameters were critical or no direction as to which of many possible choices is likely to be successful," courts must "reject hindsight claims of obviousness." Id. at 1071 (quoting In re Kubin, 561 F.3d 1351, 1359 (Fed. Cir. 2009)). Finally, evidence of obviousness based on an "obvious-to-try" theory is "insufficient unless it indicates that the possible options skilled artisans would have encountered were 'finite,' 'small,' or 'easily traversed,' and that skilled artisans would have had a reason to select the route that produced the claimed invention." Id. (quoting Ortho-McNeil Pharm., Inc. v. Mylan Labs, Inc., 520 F.3d 1358, 1364 (Fed. Cir. 2008)).

Here, the trial record reflects that a POSA would have a reasonable expectation of success in combining the prior art references and experimenting with dosing IMGN853 based on AIBW instead of TBW. The prior art indicates that a POSA faced with ocular toxicity from the administration of IMGN853 would have a "finite" number of options to try and would have a "reason to select the route that produced

the claimed invention." Id. Even though the mechanism underlying the causation of ocular toxicity by the DM4 payload was not well understood, both Dr. Figg and Dr. Tolcher testified that skilled artisans understood that the dose was relevant to toxicity,⁸⁹ and a skilled artisan would therefore have a reasonable expectation that changing the dose would impact the resulting ocular toxicity. In deciding how to alter the dose, a POSA would have a finite number of options to try: in this regard, Dr. Figg testified that the standard methods of altering the dosing for anticancer drugs in order to reduce side effects are (1) dose reduction, (2) changing the dose frequency, or (3) assessing how to calculate the dose.⁹⁰ In terms of assessing how to calculate the dose, Dr. Figg testified that there are four options other than using total body weight, all of which are disclosed in the Green reference: (1) body surface area, (2) ideal body weight, (3) lean body weight, and (4) AIBW dosing.⁹¹ Indeed, ImmunoGen's witness, Dr. Ponte, similarly testified that in considering alternative metrics for calculating body weight when developing the '809 Application, he considered body surface area, ideal body weight, lean body weight, and AIBW.⁹² And as Dr. Figg explained, AIBW is a standard tool in the toolkit for skilled artisans in the field of drug development.⁹³

89 Tr. Nov. 3, 2022 (56:19-23) (Figg); Tr. Nov. 1, 2022 (164:23-25; 165:5-8) (Tolcher).

90 Tr. Nov. 3, 2022 (40:5-7) (Figg).

91 Tr. Nov. 3, 2022 (40:12-13) (Figg).

92 Tr. Nov. 1, 2022 (60:23-25; 61:1-3) (Ponte).

93 Tr. Nov. 3, 2022 (41:19-20) (Figg) ("[I]t is common knowledge to [use AIBW] in the field of drug development of anticancer agents.").

The Alffenaar reference in the prior art provides yet another reason why a POSA would have a reasonable expectation of success in using AIBW dosing for IMGN853 to reduce ocular toxicity. The Alffenaar reference discloses that prior to the '809 Application, researchers considered using AIBW dosing as a means to reduce ocular toxicity associated with the antibiotic ethambutol. See supra ¶ 62. Although the researchers ultimately recommended using IBW instead of AIBW, the reference explains that AIBW is a method that the researchers considered specifically to address the problem of ocular toxicity. The reference therefore provides "direction as to which of many possible choices is likely to be successful," In re Cyclobenzaprine, 676 F.3d at 1071, and provides further evidence that a POSA would have a reasonable expectation of success in experimenting with AIBW dosing for IMGN853 in order to eliminate ocular toxicity.

Finally, once a POSA chose AIBW out of the limited options for calculating body weight, the 5 mg/kg and 6 mg/kg doses disclosed in Lutz '282 and the 2013 ASCO abstract would provide "parameters" of where to start with AIBW dosing. In re Cyclobenzaprine, 676 F.3d at 1071. Thus, this is not a case where ImmunoGen discovered the AIBW dosing regimen by "merely throw[ing] metaphorical darts at a board" in hopes of arriving at a successful result. Id. (quoting In re Kubin, 561 F.3d at 1359). Instead, because the prior art gave an "indication of which parameters were critical" and provided "direction as to which of many possible choices is likely to be successful," id., a POSA would have a reasonable expectation of success in experimenting with AIBW dosing.

Seeking to avoid this conclusion, ImmunoGen argues that a POSA as of October 8, 2013 would not have a reasonable expectation of success in using AIBW dosing to eliminate ocular toxicity for

IMGN853 because immunoconjugates are difficult to develop and often fail during human trials due to toxic side effects. Thus, according to ImmunoGen, no POSA could have a reasonable expectation of success without first conducting human testing, because without clinical trials, a POSA would be unable to predict how an immunoconjugate would perform in humans. This argument must be rejected because it misinterprets binding Federal Circuit precedent. The Federal Circuit has repeatedly emphasized that for a patent or patent application to be obvious, "the law does not require certainty," Eli Lilly & Co. v. Teva Pharm. Int'l GmbH, 8 F.4th 1331, 1348 (Fed. Cir. 2021), nor does it require "absolute predictability of success." OSI Pharm. v. Apotex Inc., 939 F.3d 1375, 1385 (Fed. Cir. 2019). And the Federal Circuit has clearly stated that "efficacy data," though helpful, is not "always required for a reasonable expectation of success." Id. If ImmunoGen's theory of reasonable expectation of success were adopted, any dosing regimen would be patentable, because as ImmunoGen's expert Dr. Tolcher explained, all doses must undergo human trials in order to confirm their efficacy.⁹⁴ Thus, in conclusion, the fact that the field of immunoconjugates is unpredictable is not enough to defeat an obviousness determination where, as here, (i) the problem of ocular toxicity was known, (ii) skilled artisans understood that changing the dose was a possible solution for adverse side effects such as ocular toxicity, and (iii) the prior art disclosed AIBW dosing as a potential means to eliminate or ameliorate ocular toxicity.

94 Tr. Nov. 2, 2022 (96:2-5) (Tolcher).

6. Secondary Indicia of Obviousness

The final step in the obviousness analysis is to determine whether secondary considerations demonstrate that the '809 Application is unobvious. In Graham, the Supreme Court identified three "secondary indicia" of obviousness that courts should consider: (i) commercial success, (ii) long-felt but unresolved needs, and (iii) failure of others to find a solution to the problem or make the invention. Graham, 383 U.S. at 17, 86 S.Ct. 684. Since then, the Federal Circuit has identified additional secondary considerations to be considered indicia of obviousness, including copying by competitors, unexpected results, and industry praise.⁹⁵ The Federal Circuit has explained that because objective indicia of non-obviousness are "crucial in avoiding the trap of hindsight" when reviewing what otherwise seems like "a combination of known elements," Leo Pharm. Prods., 726 F.3d at 1358, secondary objective criteria "must always be considered and given whatever weight is warranted by the evidence presented." Knoll Pharm. Co., Inc. v. Teva Pharm. USA, Inc., 367 F.3d 1381, 1385 (Fed. Cir. 2004) (per curiam). Furthermore, Federal Circuit precedent makes clear that secondary considerations such as unexpected results may arise after the filing date. Genetics Inst., LLC v. Novartis Vaccines & Diagnostics, Inc., 655 F.3d 1291, 1308 (Fed. Cir. 2011). Thus, even though the § 103 analysis "remains properly focused 'at the time the invention was made,'" a patent applicant may present "relevant indicia of nonobviousness, whether or not this evidence was available or expressly contemplated at the filing of the patent application." Id. Finally, "in order to accord

95 See, e.g., Liqwd, Inc. v. L'Oreal USA, Inc., 941 F.3d 1133, 1137 (Fed. Cir. 2019) (copying); Millennium Pharms., Inc. v. Sandoz Inc., 862 F.3d 1356, 1368 (Fed. Cir 2017) (unexpected results); Apple Inc. v. Samsung Elec. Co., Ltd., 839 F.3d 1034, 1053 (Fed. Cir. 2016) (en banc) (industry praise).

substantial weight to secondary considerations, the evidence of secondary considerations must have a 'nexus' to the claims, i.e., there must be a legally and factually sufficient connection between the evidence and the patented invention." Teva Pharm. Int'l GmbH v. Eli Lilly & Co., 8 F.4th 1349, 1360 (Fed. Cir. 2021) (quoting Henny Penny Corp. v. Frymaster LLC, 938 F.3d 1324, 1332 (Fed. Cir. 2019)). And importantly, the patent applicant "bears the burden of showing that a nexus exists." Id.

ImmunoGen argues that unexpected results, long-felt need, failure of others, copying, and industry praise demonstrate that the '809 Application is not obvious. In response, the PTO argues that the secondary indicia do not support a finding of unobviousness. Each of the disputed secondary indicia of obviousness is discussed in detail below, but in short, none of the secondary factors persuasively demonstrates that the '809 Application is unobvious.

i. Unexpected Results

To begin with, ImmunoGen argues that the AIBW dosing regimen described in the '809 Application unexpectedly resulted in eliminating the ocular toxicity previously associated with IMGN853. According to ImmunoGen, the reduction in severe adverse ocular events following the switch to AIBW dosing would have been unexpected to a skilled artisan, and without this reduction in adverse events, IMGN853 would not have advanced beyond the initial Phase I trials. In response, the PTO argues that ImmunoGen has failed to show the existence of unexpected results because clinical trial results do not in fact confirm that AIBW dosing reduces ocular toxicity relative to dosing based on TBW.

The Federal Circuit has explained that unexpected results are objective evidence of unobviousness because "that which would have been surprising to a person of ordinary skill in a particular art would not have been obvious." In re Soni, 54 F.3d 746, 750 (Fed. Cir. 1995). It is also well settled, however that unexpected results must be supported with "objective evidentiary support" and "factual evidence." CFMT, Inc. v. Yieldup Int'l Corp., 349 F.3d 1333, 1342 (Fed. Cir. 2003). Here, ImmunoGen, which has the burden of proving secondary considerations, has failed to demonstrate by a preponderance of the evidence that the AIBW dosing in the '809 Application led to the unexpected result of reducing or eliminating ocular toxicity for IMGN853. Instead, there is insufficient evidence in the record to conclude that AIBW dosing in fact reduces or eliminates ocular toxicity.

In attempting to demonstrate unexpected results, ImmunoGen relies on the early Phase I study of IMGN853, which disclosed no incidents of ocular toxicity in the small group of individuals who initially received IMGN853 dosed according to AIBW. In making this argument, ImmunoGen ignores more recent studies of AIBW dosing for IMGN853, which call into question the early Phase I results. Because district courts are permitted to look to evidence of secondary considerations that arises after the priority date of the application at issue, see Genetics Inst., 655 F.3d at 1308, those later studies are pertinent to consider here. Specifically, as described in detail in the Findings of Fact above, see supra ¶¶ 87-96, the Forward 1 study and the Soraya study both undermine the Phase I results and demonstrate that even with AIBW dosing, ocular toxicity remains a problem for patients given IMGN853.

First, the Forward 1 study indicates that some patients who received IMGN853 at a dose of 6 mg/kg AIBW experienced Grade 3 ocular toxicity even after receiving prophylactic treatments such as lubricating eye drops. Furthermore, during the

course of the Forward 1 study, ImmunoGen excluded patients who had active or chronic eye problems. Because of these measures, Dr. Figg explained that the ocular toxicity results reported in the Forward 1 study are "falsely lower" than they would have been had ImmunoGen applied the same protocols that it did during the Phase I studies.⁹⁶ Dr. Figg also explained that despite the added prophylactic measures that ImmunoGen took during the Forward 1 study, ImmunoGen's more recent publications show that 20% of the patients in the Forward 1 study needed either delayed dosing or dose reductions because the patients experienced ocular toxicity. As Dr. Figg testified, this confirms that ocular toxicity remains a serious problem for IMGN853, even with AIBW dosing.⁹⁷

96 Tr. Nov. 3, 2022 (69:8) (Figg).

97 Tr. Nov. 3, 2022 (77:11-25) (Figg).

Second, the Soraya study also undermines the Phase I results for AIBW dosing of IMGN853. The Soraya study included all the same prophylactic measures designed to reduce ocular toxicity as the Forward 1 study. Despite this, 12 patients, or 11%, experienced Grade 3 ocular toxicity. Thus, the Soraya study does not confirm ImmunoGen's assertion that AIBW dosing unexpectedly eliminates ocular toxicity in patients receiving IMGN853.

Seeking to avoid this conclusion, ImmunoGen relies on the 2022 ESMO poster which reports results of a covariate analysis conducted using the data from ImmunoGen's Phase I study, the Forward 1 study, and the Soraya study. According to ImmunoGen, the 2022 ESMO poster indicates that AIBW dosing is safe and effective for all patients regardless of weight. Specifically, the ESMO poster reports that AIBW was not a covariate that impacted the efficacy of IMGN853 and its dosing regimen; Dr. Shah testified that this means dosing IMGN853 at 6 mg/kg AIBW results in lower incidences of toxicity for all patients, regardless of weight.⁹⁸ This argument is unpersuasive. First, the record reflects that the ESMO poster deserves little weight here. The ESMO poster was not disclosed in the ordinary course of discovery, and the experts had little time to review the document before testifying about it during the bench trial. Furthermore, the ESMO poster reports the results of a covariate analysis, but precious little testimony was presented by either side about the methodology or reliability of covariate studies, and no evidence was presented about the underlying assumptions in the covariate analysis reported in the ESMO poster. Thus, on this record, there is not enough information to determine whether AIBW dosing is safe and effective for patients of all weights. And as Dr. Figg explained, AIBW dosing may in fact be dangerous for underweight patients, because patients who weigh less than their ideal body weight receive a higher dose of IMGN853 when dosed based on AIBW than they would receive if dosed based on TBW. See supra ¶ 95. Cachectic or underweight patients therefore receive a larger amount of the toxic DM4 payload under the claimed AIBW dose for IMGN853 than they otherwise would. Id. Because of this, Dr. Figg explained that the persisting incidents of ocular toxicity in the Forward 1 and Soraya studies may have occurred in underweight patients who received too much of the toxic payload. See supra ¶ 96. Of course, further studies in the future may reveal that AIBW dosing of IMGN853 is indeed therapeutically effective and reduces ocular toxicity for patients across the weight spectrum, but on this record, ImmunoGen has failed to provide information

98 Tr. Nov. 3, 2022 (170:23-25) (Shah).

that by a preponderance of the evidence demonstrates that AIBW dosing led to the unexpected effect of reducing ocular toxicity while maintaining therapeutic efficacy for all patients.

In conclusion, ImmunoGen has failed to demonstrate that the AIBW dosing regimen claimed in the '809 Application in fact reduces or eliminates ocular toxicity. The secondary factor of unexpected results therefore does not weigh in favor of finding the '809 Application unobvious.

ii. Long-Felt, Unmet Need

ImmunoGen next argues that the AIBW dosing regimen claimed in the '809 Application met a decades-long need for a tolerable and effective platinum-resistant ovarian cancer treatment. As with unexpected results, this argument fails to persuade. Federal Circuit precedent teaches that "[e]vidence of a long felt but unresolved need tends to show non-obviousness because it is reasonable to infer that the need would have not persisted had the solution been obvious." WBIP, LLC v. Kohler Co., 829 F.3d 1317, 1332 (Fed. Cir. 2016). But as with any secondary obviousness consideration, the asserted long-felt need must have a nexus to the claimed invention. Eli Lilly & Co., 8 F.4th at 1360. Finally, whether there is a "long-felt need is analyzed as of the date of an articulated identified problem and evidence of efforts to solve that problem." WBIP, 829 F.3d at 1334 (quoting Tex. Instruments, Inc. v. U.S. Int'l Trade Comm'n, 988 F.2d 1165, 1178 (Fed. Cir. 1993)). Here, a long-felt need does not weigh in favor of finding the '809 Application unobvious for the simple reason that ImmunoGen cannot show that such a need existed as of October 8, 2013.

ImmunoGen argues that the '809 Application solves the long-felt need for a treatment for platinum-resistant ovarian cancer. But this misinterprets the requisite analysis. The '809 Application does not seek to patent the drug IMGN853, because ImmunoGen already has multiple patents for IMGN853. Instead, the '809 Application is directed to a dosing method for administering IMGN853 in order to eliminate ocular toxicity. Thus, the relevant need that ImmunoGen must show is satisfied by the '809 Application is a need for a dosing regimen for IMGN853 that solves the problem of ocular toxicity. The ocular toxicity problems associated with the administration of IMGN853 were first observed during Phase I testing in 2012, and were not publicly disclosed in 2017. Thus, as of the priority date of the '809 Application on October 8, 2013, there was no publicly recognized, long-felt need for AIBW dosing for IMGN853 at all.

Seeking to avoid this conclusion, ImmunoGen points to two Federal Circuit cases holding that a long-felt need for a safe and tolerable treatment where existing treatments had negative side effects supported the unobviousness of a patent. See Forest Labs., 918 F.3d at 936-37; Procter & Gamble Co. v. Teva Pharms. USA, Inc., 566 F.3d 989, 998 (Fed. Cir. 2009). But neither Forest Laboratories nor Procter & Gamble is apposite here. First, in Forest Laboratories, the Federal Circuit noted that demonstrating that a claimed drug reduced negative side effects "may not be a particularly strong demonstration of a long-felt need." 918 F.3d at 936-37 (emphasis added). Despite this, because the existence of a long-felt need is a question of fact, the Federal Circuit affirmed the district court's conclusion that the long-felt need weighed in favor of unobviousness because the district court "did not clearly err[]." Id. at 937. Thus, Forest Laboratories is hardly strong support for ImmunoGen's position. At best, it holds that the district court's conclusion in that case was not clear error, but demonstrates that reduction of negative side effects may not be strong evidence of a long-felt need. In this

case, the evidence was insufficient for ImmunoGen to carry its burden and demonstrate that the invention claimed in the '809 Application satisfies a long-felt need. Nor is Procter & Gamble any help to ImmunoGen. There, the Federal Circuit affirmed the district court's finding that an osteoporosis treatment satisfied a long-felt need because at the time of the filing date, all other treatments for osteoporosis were "inadequate." Procter & Gamble, 566 F.3d at 998. But unlike this case, in Procter & Gamble, the claimed invention was the drug itself, not a dosing regimen. Thus, the district court concluded that the claimed drug solved an unmet need because it more effectively treated osteoporosis than other drugs on the market. Here, the drug—IMGN853—is already patented and no evidence at trial showed that reducing ocular toxicity caused by IMGN853 was a long-felt need.⁹⁹ Furthermore, in Forest Laboratories and Procter & Gamble, the district court determined that the claimed invention actually addressed the asserted need. Here, as discussed at length above, ImmunoGen has failed to show that AIBW dosing in fact reduces ocular toxicity when compared to TBW dosing. Thus, Forest Laboratories and Procter & Gamble are inapposite and unpersuasive. In conclusion, ImmunoGen has failed to show that a long-felt need supports the unobviousness of the '809 Application, because as of the priority date of the '809 Application, there was no publicly recognized long-felt need for an AIBW dosing regimen for IMGN853.

99 There is no doubt that anticancer drugs are a long-felt need in other circumstances, as there will always be a need for drugs to treat serious illnesses. But because the anticancer drug described in the '809 Application has already been patented, see supra ¶¶ 102-07, the invention claimed in the '809 Application is limited to the AIBW dosing regimen for IMGN853. Thus, because secondary considerations must have a "nexus" to the "patented invention," Eli Lilly & Co., 8 F.4th at 1360, the long-felt need must be tailored to the dosing regimen for IMGN853. Here, ImmunoGen has only demonstrated a long-felt need for anticancer drugs and has failed to carry its burden to show persuasively a nexus to the dosing regimen claimed in the '809 Application.

Finally, ImmunoGen points to the FDA's "Fast Track" approval of IMGN853 as evidence that the '809 Application meets the need for a safe and tolerable treatment for platinum-resistant ovarian cancer. But the FDA's grant of accelerated approval for IMGN853 does not confirm that the '809 Application meets a long-felt need. Instead, the FDA approval requires ImmunoGen to conduct additional studies of the dosing regimen for IMGN853 in order to "characterize the safety of the recommended dosage of 6 mg/kg AIBW."¹⁰⁰ This indicates that although the FDA views IMGN853 as valuable, the FDA seeks additional data in order to confirm whether the 6 mg/kg AIBW dosing indeed works better than other dosing methods. Thus, in conclusion, ImmunoGen has failed to carry its burden to show that IMGN853 satisfies the long-felt, unmet need for a dosing regimen for IMGN853, and the FDA "Fast Track" approval of IMGN853 does not change this result. The secondary consideration of a long-felt but unmet need therefore does not support a finding that the '809 Application is unobvious.

100 FDA Approval Letter at 5.

iii. Failure of Others

ImmunoGen also contends that failure of others to address toxicities associated with immunoconjugates supports the conclusion that the '809 application is unobvious. The Federal Circuit has explained that "[n]onobviousness is suggested by the failure of others to find a solution to the problem which the patent[] in question purport[s] to solve." Symbol

Techs., Inc. v. Opticon, Inc., 935 F.2d 1569, 1578-79 (Fed. Cir. 1991). At trial, ImmunoGen presented brief testimony from Dr. Tolcher that ImmunoGen's competitors, including Sanofi, Bayer, AbbVie, and Seattle Genetics have failed to develop immunoconjugates because they were not able to overcome toxic side effects.¹⁰¹ This does not support the conclusion that the '809 Application is unobvious, however, because ImmunoGen is not seeking to claim a method of dosing all immunoconjugates. Instead, the '809 Application is narrowly directed at a dosing regimen for IMGN853 in order to reduce ocular toxicity. ImmunoGen has presented no evidence that any other drug developers have tried and failed to solve the same problem that the '809 Application purports to solve: that is, eliminating ocular toxicity associated with IMGN853.¹⁰² Thus, failure of others does not persuasively demonstrate that the '809 Application is unobvious.

101 Tr. Nov. 1, 2022 (157:7-12; 161:19-25) (Tolcher).

102 Indeed, because only ImmunoGen could profit from the use of IMGN853, no other entity would have an incentive to find a dosing regimen for IMGN853 that reduced ocular toxicity. See supra note 67.

iv. Copying

ImmunoGen next argues that copying indicates that the '809 Application is unobvious. At trial, ImmunoGen presented testimony from Dr. Tolcher that Sutro Biopharma, one of ImmunoGen's competitors, adopted AIBW dosing in clinical trials for an immunoconjugate that treats ovarian cancer.¹⁰³ But this testimony does not support the conclusion that the '809 Application is unobvious. As the Federal Circuit has explained, "[a] showing of copying is only equivocal evidence of non-obviousness in the absence of more compelling objective indicia of other secondary considerations." Sealy Tech., LLC v. SSB Manuf. Co., 825 F. App'x 801, 808 (Fed. Cir. 2020) (quoting Ecolochem, Inc. v. S. Cal. Edison Co., 227 F.3d 1361, 1380 (Fed. Cir. 2000)). Moreover, the Federal Circuit has explained that evidence of copying is entitled to "little weight" if the copier's "product is not identical to the claimed invention." Pentec, Inc. v. Graphic Controls Corp., 776 F.2d 309, 317 (Fed. Cir. 1985). ImmunoGen introduced no evidence that ImmunoGen's competitor used AIBW dosing for an immunoconjugate identical to IMGN853. Indeed, Dr. Tolcher agreed that the immunoconjugate developed by Sutro was a "different drug."¹⁰⁴ Thus, ImmunoGen has not satisfied its burden to demonstrate that copying by competitors supports finding the '809 Application unobvious.

103 Tr. Nov. 1, 2022 (209:8-210:15) (Tolcher).

104 Tr. Nov. 1, 2022 (210:19-21) (Tolcher).

v. Industry Praise

Finally, ImmunoGen argues that IMGN853 has been praised, citing a trade publication and an investment analyst sheet from Canaccord Genuity. Evidence of "industry praise of the claimed invention ... weighs in favor of the non-obviousness of the claimed invention." Apple, 839 F.3d at 1053. But here, there is no industry praise in the record of the claimed invention; all of ImmunoGen's cited industry praise pertains only to the anticancer properties of IMGN853, for which ImmunoGen has already received multiple patents. As Dr. Figg made clear, none of the industry praise pertains to the claimed AIBW dosing method in the '809 Application—indeed, none of the cited praise mentions AIBW at all.¹⁰⁵ Thus, ImmunoGen

105 Tr. Nov. 3, 2022 (85:1-5) (Figg).

has failed to show a nexus between the praise and the claimed invention, and industry praise therefore cannot support a finding of unobviousness here.

In conclusion, ImmunoGen has failed to demonstrate that secondary considerations support finding that the '809 Application is unobvious. Thus, each step of the Graham analysis leads to the conclusion that the '809 Application is invalid as obvious. Put simply, the trial record persuasively reflects that (i) the elements of the '809 Application were disclosed in the prior art, (ii) a POSA would have been motivated to combine the prior art references to arrive at the claimed AIBW dosing method, and (iii) a skilled artisan would have a reasonable expectation of success in doing so. The patent examiner and the PTAB therefore correctly determined that the '809 Application is obvious in violation of § 103.

C. Obviousness-Type Double Patenting

The final issue is whether the '809 Application is invalid under the doctrine of obviousness-type double patenting, as both the PTO and the PTAB concluded. Obviousness-type double patenting is a judicially created doctrine grounded in § 101 of the Patent Act. Abbvie, 764 F.3d at 1372. Section 101 provides that "[w]hoever invents or discovers any new and useful process, machine, manufacture, or composition of matter ... may obtain a patent therefor." 35 U.S.C. § 101. The Federal Circuit has recognized that § 101 forbids an individual from obtaining more than one patent on the same invention, which is known as "double patenting." Abbvie, 764 F.3d at 1372. Thus, under this doctrine, an inventor cannot "secur[e] a second, later expiring patent for the same invention." Id. at 1373. The prohibition on receiving two patents for the same invention has existed since the inception of our patent laws: as early as 1819, Justice Story explained that "if [a patentee] can successively take out at different times new patents for the same invention, he may perpetuate his exclusive right during a century." Id. (quoting Odiorne v. Amesbury Nail Factory, 18 F. Cas. 578, 579 (C.C.D. Mass. 1819)). The ban on double patenting therefore "ensures that the public gets the benefit of the invention after the original period of monopoly expires." Id. For this reason, under the obviousness-type double patenting doctrine, a later claim that is obvious over or anticipated by an earlier claim is invalid for obviousness-type double patenting. Id. at 1374.

Here, the '809 Application clearly fails under the doctrine of obviousness-type double patenting. As previously described, ImmunoGen has already received four patents covering IMGN853 and its use for treating cancer: the '966 patent, the '275 patent, the '490 patent, and the '280 patent. See supra ¶¶ 102-07. The relevant question is therefore whether the claims in the '809 Application would have been obvious over the issued claims in the four earlier patents. Abbvie, 764 F.3d at 1374. As previously addressed in detail, the '809 Application is fatally obvious because AIBW dosing was disclosed in the prior art, and a POSA would have both a motivation to combine the prior art references and a reasonable expectation of success in doing so.¹⁰⁶ Allowing ImmunoGen to receive another patent for the '809 Application

106 The parties both agree that the issue of obviousness-type double patenting rises or falls with the obviousness determination. See Pl.'s Trial Brief, Dkt. 117 at 18 ("The Office's obviousness-type double patenting contentions rise or fall with its obviousness grounds"); Tr. Nov. 1, 2022 (42:16-17) (Counsel for the PTO stated that "the obvious type double patenting ... rises and falls with obviousness.").

would therefore permit ImmunoGen to extend its patent term by obtaining a second, later expiring patent for the same invention—which is exactly what Justice Story warned against. In other words, because the '809 Application is fatally obvious, it must also fail for obviousness-type double patenting. Accordingly, the doctrine of obviousness-type double patenting provides a third reason why a patent cannot issue for the '809 Application.

* * *

In conclusion, the PTO and the PTAB properly determined that ImmunoGen is not entitled to a patent for the AIBW dosing regimen claimed in the '809 Application. The '809 Application is fatally indefinite and obvious; furthermore, the '809 Application is invalid under the doctrine of obviousness-type double patenting. Accordingly, judgment must enter in favor of the PTO.

An appropriate Order will issue.