On behalf of: Par Pharmaceutical, Inc.
`
`
`
`
`
`
`
`
`
`Entered: August 3, 2017
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`_______________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`_______________________
`
`PAR PHARMACEUTICAL, INC.
`Petitioner
`
`v.
`
`NOVARTIS AG
`Patent Owner
`_______________________
`
`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`_______________________
`
`Before LORA M. GREEN, CHRISTOPHER L. CRUMBLEY, and
`ROBERT A. POLLOCK, Administrative Patent Judges.
`
`
`PETITIONER’S REPLY
`
`
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`
`TABLE OF CONTENTS
`
`I.
`
`II.
`
`Introduction ...................................................................................................... 1
`
`The Claimed Methods Would Have Been Obvious ........................................ 2
`
`C.
`
`D.
`
`E.
`
`F.
`
`A. Dr. Ratain Qualifies as a POSA ............................................................ 2
`B.
`Clinicians Routinely Administered the Same Treatments for
`PNET and Carcinoid Tumors ................................................................ 4
`The Teachings of Oberg 2004 Are Relevant to a POSA Seeking
`to Develop New PNET Treatments ....................................................... 7
`The ’224 Patent and Prior Art Group Rapamycin, Everolimus,
`and Temsirolimus as a Class of Anticancer Agents .............................. 8
`A POSA Would Not Have Expected PNETs “After Failure of
`Cytotoxic Chemotherapy” to Be “More Resistant” ............................ 10
`A POSA Would Have Reasonably Expected that Everolimus
`Would Be Therapeutically Effective in Advanced PNETs ................. 13
`1.
`The role of mTOR inhibitors as anticancer agents,
`including in NETs, was not “uncertain” ................................... 13
`A POSA would consider preclinical models for
`somatostatin receptor-positive tumors as models for
`NETs ......................................................................................... 17
`Oberg 2004 and Duran provide a reasonable expectation
`that everolimus would be therapeutically effective in
`PNETs ....................................................................................... 18
`
`2.
`
`3.
`
`III. Novartis Has Not Established Sufficient Evidence of Secondary
`Indicia to Overcome the Strong Showing of Obviousness............................ 22
`
`A.
`
`2.
`
`Everolimus Does Not Have Any Unexpected Results Compared
`to the Closest Prior Art ........................................................................ 22
`1.
`Novartis presents no difference between use of
`everolimus and rapamycin or temsirolimus to treat NETs ....... 22
`Novartis presents no difference between use of
`everolimus as monotherapy and in combination with a
`somatostatin .............................................................................. 23
`Novartis presents no difference between use of
`everolimus after failure of cytotoxic chemotherapy and
`chemotherapy-naïve NETs ........................................................ 24
`FDA-approval status is not a property of everolimus ............... 24
`
`3.
`
`4.
`
`i
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`
`B.
`
`Everolimus Did Not Satisfy Any Long-Felt But Unmet Needs
`or Succeed Where Others Had Failed ................................................. 25
`
`IV. Conclusion ..................................................................................................... 25
`
`
`
`
`
`ii
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`
`TABLE OF AUTHORITIES
`
`CASES
`
`Page(s)
`
`Alcon Research, Ltd. v. Apotex Inc.,
`687 F.3d 1362 (Fed. Cir. 2012) .......................................................................... 13
`
`Allergan, Inc. v. Apotex Inc.,
`754 F.3d 952 (Fed. Cir. 2014) ..................................................................... 14, 15
`
`Bristol-Myers Squibb Co. v. Teva Pharm. USA, Inc.,
`752 F.3d 967 (Fed. Cir. 2014) ............................................................................ 22
`
`CSS Fitness, Inc. v. Brunswick Corp.,
`288 F.3d 1359 (Fed. Cir. 2002) ............................................................................ 3
`
`Daubert v. Merrell Dow Pharm., Inc.,
`509 U.S. 579 (1993) ............................................................................................ 21
`
`Genzyme Therapeutic Prods. L.P. v. Biomarin Pharm. Inc.,
`825 F.3d 1360 (Fed. Cir. 2016) .......................................................................... 15
`
`In re Mouttet,
`686 F.3d 1322 (Fed. Cir. 2012) ............................................................................ 8
`
`In re O’Farrell,
`853 F.2d 894 (Fed. Cir. 1988) ............................................................................ 13
`
`Merck & Cie v. Gnosis SpA,
`808 F.3d 829 (Fed. Cir 2015) ............................................................................. 23
`
`Merck & Co. v. Teva Pharm. USA, Inc.,
`395 F.3d 1364 (Fed. Cir. 2005) ................................................................... 12, 19
`
`Ormco Corp. v. Align Tech., Inc.,
`463 F.3d 1299 (Fed. Cir. 2006) .......................................................................... 24
`
`Pfizer, Inc. v. Apotex, Inc.,
`480 F.3d 1348 (Fed. Cir. 2007) .......................................................................... 13
`
`iii
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`PharmaStem Therapeutics, Inc. v. ViaCell, Inc.,
`491 F.3d 1342 (Fed. Cir. 2007) ..................................................................... 9, 10
`
`Phillips v. AWH Corp.,
`415 F.3d 1303 (Fed. Cir. 2005) ............................................................................ 3
`
`SEB S.A. v. Montgomery Ward & Co.,
`594 F.3d 1360 (Fed. Cir. 2010) ............................................................................ 3
`
`Smith & Nephew, Inc. v. Rea,
`721 F.3d 1371 (Fed. Cir. 2013) ............................................................................ 8
`
`Toro Co. v. MTD Prods. Inc.,
`IPR2016-00219, Paper 39 (P.T.A.B May 10, 2017) ............................................ 3
`
`OTHER AUTHORITIES
`
`M.P.E.P. § 2107.03(IV) (2015) ................................................................................ 20
`
`
`
`iv
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`I.
`
`INTRODUCTION
`
`Par presented compelling evidence in its petition that the ’224 patent claims
`
`were obvious. Specifically, Dr. Mark Ratain’s testimony demonstrated that two
`
`different, independent prior-art researchers had published the use of mTOR
`
`inhibitors to treat NETs after the failure of cytotoxic chemotherapy: Oberg 2004
`
`(rapamycin) and Duran (temsirolimus). The only difference between the ’224
`
`patent claims and these prior-art disclosures is that the claims specify a closely-
`
`related mTOR inhibitor, everolimus, used to treat PNETs, a type of NET.
`
`As demonstrated by Dr. Ratain’s testimony, the prior art would have
`
`motivated a POSA to use everolimus to treat PNETs after failure of cytotoxic
`
`chemotherapy, and provided a reasonable expectation of success. The prior art
`
`uniformly grouped everolimus, temsirolimus, and rapamycin as a class of mTOR
`
`inhibitors based on their common mechanism of action. The prior art also
`
`disclosed that everolimus had antitumor activity in a model for PNETs and
`
`reported clinical efficacy in solid tumors. Accordingly, the prior art provides the
`
`motivation and the reasonable expectation of success rendering the claims obvious.
`
`Novartis’s response attempts to distract from this clear showing with
`
`evidence that is not credible, and generally not legally relevant. Novartis first
`
`attacks Dr. Ratain’s expertise. The evidence shows that Dr. Ratain is a highly-
`
`qualified expert in oncology drug development. Novartis places great weight on its
`
`1
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`assertion that tumors that had failed to respond to chemotherapy would be
`
`considered difficult to treat and less likely to respond to treatment. Novartis
`
`cannot criticize the prior art on this point as the ’224 patent provides no data
`
`evidencing everolimus is effective in these tumors. Nevertheless, Novartis’s
`
`assertion is simply wrong and unsupported by Novartis’s cited literature. The
`
`literature overwhelmingly shows that researchers understood that prior failure of
`
`chemotherapy had no effect on the expected efficacy of a molecularly targeted
`
`therapy, such as an mTOR inhibitor. Novartis’s remaining arguments all suffer
`
`from the fundamental legal error of substituting a “guarantee” of success for the
`
`required “reasonable expectation.” Novartis points to the supposed “uncertainty”
`
`in PNET etiology, lack of conclusive clinical data, and supposed differences
`
`between everolimus and rapamycin and temsirolimus, or PNETs compared to
`
`NETs generally. Finally, Novartis asserts secondary considerations allegedly
`
`support the claims’ non-obviousness. Each of Novartis’s arguments is based on
`
`inaccurate, inconsistent, incomplete, or irrelevant analyses and fails to overcome
`
`the clear disclosures in the prior art.
`
`II. THE CLAIMED METHODS WOULD HAVE BEEN OBVIOUS
`A. Dr. Ratain Qualifies as a POSA
`Novartis first alleges that Dr. Ratain would not qualify as a POSA and his
`
`testimony “should be given little weight.” (POR 4-7.) Novartis’s arbitrary
`
`2
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`standard, specifically seeking to exclude Dr. Ratain, cannot be correct, as it
`
`appears to exclude the ’224 patent’s inventors. (Ex. 1119, Ratain Supp. Decl. ¶¶7-
`
`16.) Phillips v. AWH Corp., 415 F.3d 1303, 1313 (Fed. Cir. 2005) (en banc)
`
`(stating the “well-settled understanding that inventors are typically persons skilled
`
`in the field of the invention”); CSS Fitness, Inc. v. Brunswick Corp., 288 F.3d
`
`1359, 1368 (Fed. Cir. 2002) (same). The ’224 patent is directed to oncology drug
`
`development, a field in which Dr. Ratain—like named inventor Dr. Lebwohl—has
`
`decades of expertise. (Ex. 1119, Ratain Supp. Decl. ¶¶7-15.)
`
`In any event, Dr. Ratain easily meets the requirements of Novartis’s
`
`proposed standard. Dr. Ratain has the experience Novartis’s expert identifies a
`
`“specialist” in NETs would possess, including the use of mTOR inhibitors for
`
`treating NETs. (Id. ¶¶11-12; Ex. 1070, Kulke Dep. 51:8-18.) Thus, Dr. Ratain
`
`was a POSA as of November 2005 under either party’s definition.
`
`Further, Dr. Ratain is an expert in medical oncology, and an expert witness
`
`is not required to ever have been a POSA. SEB S.A. v. Montgomery Ward & Co.,
`
`594 F.3d 1360, 1372-73 (Fed. Cir. 2010); Toro Co. v. MTD Prods. Inc., IPR2016-
`
`00219, Paper 39 at 6 (P.T.A.B. May 10, 2017). Dr. Ratain’s testimony is
`
`completely within his field of expertise.
`
`3
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`
`B. Clinicians Routinely Administered the Same Treatments for
`PNET and Carcinoid Tumors
`
`Oberg 2004 and Duran each explicitly taught administering an mTOR
`
`inhibitor to treat NETs. (Pet. 29-32, 36-37, 40-46, 49-51.) Novartis nevertheless
`
`argues that because these clear teachings relate to the treatment of NETs
`
`“generally,” they do not apply to the “specific” subset of PNETs.1 (POR 7-12, 26-
`
`27.) Novartis argues that because NETs are “a heterogeneous group of tumors,” a
`
`POSA “would not have reasonably expected carcinoids and PNETs to respond in
`
`the same way to a new therapy.” (Id. 26-27.) But the disclosures in the art are
`
`clear. A POSA would have reasonably expected everolimus, a known mTOR
`
`inhibitor, to be an effective treatment for PNET based on the explicit teachings in
`
`both Oberg 2004 and Duran using mTOR inhibitors to treat NETs generally,
`
`including both PNETs and carcinoids. (Pet. 40-46, 49-51.)
`
`Further, the prior art, including the contemporaneous activities of Novartis’s
`
`1 Novartis, relying on Dr. Kulke’s declaration, includes a chart purportedly defin-
`
`ing separate categories of NETs (POR 10; Ex. 2041 ¶48), but Dr. Kulke admitted
`
`that the chart was generated by Novartis’s counsel and testified that he could not
`
`recall using such a chart to organize NETs in any paper or presentation. (Ex. 1070,
`
`Kulke Dep. 53:21-54:23.) Dr. Kulke further admitted that the separate boxes in the
`
`chart do not reflect a difference in treatment of those tumors. (See id. 56:8-57:10.)
`
`4
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`own expert, indicate that the same treatments were identified as treatments for both
`
`PNETs and carcinoids, contradicting Novartis’s argument. Oberg 2004 identifies
`
`therapy for NETs
`
`to
`
`include somatostatin analogs,
`
`interferon, cytotoxic
`
`chemotherapy, and radiolabeled DOTA-ocretotide. (Ex. 1027 Fig. 1.) Each of
`
`these are identically identified in Oberg & Eriksson’s prior-art algorithm for
`
`treating PNETs, which Novartis endorses as more “specific” to PNETs. (POR 26-
`
`27; Ex. 1028 Fig. 2.) Thus, Dr. Oberg, an undisputed expert in the treatment of
`
`NETs, (Ex. 1070, Kulke Dep. 43:8-10), taught the same treatments for both
`
`carcinoids and PNETs. Further, Dr. Kvols, Novartis’s PNET expert in the related
`
`district court litigation, explained that “APUD cells are thought to be the origin of
`
`tumors such as carcinoids. . . [and] islet cell carcinoma [PNETs] . . . . This
`
`cytologic basis of cell similarity has enabled investigators to use similar treatment
`
`programs for tumors of great anatomic diversity.” (Ex. 1096, Kvols 1987 77
`
`(emphasis added).) Finally, Dr. Kulke conceded that benign carcinoid and benign
`
`PNET are treated in the same way, and localized NETs, which includes both
`
`carcinoids and PNETs, are all treated through surgery. (Ex. 1070, Kulke Dep.
`
`56:8-57:10.)
`
`As of November 2005, Dr. Kulke regularly conducted clinical trials for new
`
`treatments for NETs that enrolled both carcinoid and PNET patients for the same
`
`clinical protocol. (Ex. 2042, Kulke CV 4-6 (identifying seven clinical trials in
`
`5
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`patients having NETs, which includes both carcinoid and PNET); Ex. 1070 Kulke
`
`Dep. 8:23-9:16, 10:24-12:12; Ex. 2014; Ex. 2099.) Thus, although PNETs and
`
`carcinoids may have different specific responses to a treatment, as of November
`
`2005, Dr. Kulke administered the same treatments to both tumor types when
`
`investigating new therapies.
`
`Further, the response rates to treatments for PNETs frequently were greater
`
`than that for carcinoids, as Novartis and Dr. Kulke admit. (POR 9 (describing
`
`PNETs as “more sensitive”); Ex. 1070, Kulke Dep. 24:7-12.) PNETs also had
`
`better responses to molecularly targeted therapies, a category that includes mTOR
`
`inhibitors but not cytotoxic chemotherapy. (Ex. 1070, Kulke Dep. 15:6-25, 158:8-
`
`15.) In 2005, Dr. Kulke identified that PNETs had better response to sunitinib (a
`
`molecularly targeted therapy) than did carcinoids. (Ex. 2099 718Ab (reporting
`
`13% partial response rate in PNET and 2% partial response rate in carcinoid).)
`
`Consistent with these data, Dr. Kulke has also explained that “most therapeutic
`
`agents demonstrat[e] higher response rates in [PNETs] compared with carcinoid.”
`
`(Ex. 1076 at 365.)
`
`Thus, if a POSA were to form any different expectation regarding the
`
`response of PNET as opposed to a carcinoid treated with an mTOR inhibitor, the
`
`prior art (including from Dr. Kulke) shows that a POSA would reasonably expect a
`
`better response from PNET. Simply put, Novartis fails to present a good reason
`
`6
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`why a POSA would have reasonably expected that an mTOR inhibitor, including
`
`everolimus, would not have antitumor activity in NETs, including PNETs, based
`
`on the teaching of Duran and Oberg 2004. (Pet. 40-46, 49-51.)
`
`C. The Teachings of Oberg 2004 Are Relevant to a POSA Seeking to
`Develop New PNET Treatments
`
`Unable to avoid the explicit teaching of Oberg 2004, Novartis attempts to
`
`discredit the reference by pointing to a separate treatment algorithm later published
`
`by Dr. Oberg. (POR 27.) According to Novartis, because this later algorithm is
`
`“specific” to PNETs and does not include rapamycin, “a POSA would not have
`
`been motivated to follow the teaching in Oberg 2004” for treating PNETs. (Id.,
`
`citing Ex. 1028; Ex. 1070, Kulke Dep. 126:18-127:11.) But unlike Oberg 2004,
`
`the “specific” PNET reference, Oberg & Ericksson, describes only those
`
`treatments already in clinical use.2 (Ex. 1028 at 767-75; Ex. 1070, Kulke Dep.
`
`136:14-138:8.) Oberg & Eriksson explicitly states that “[n]ew therapies based on
`
`current knowledge of tumour biology are warranted.” (Ex. 1028 at 775.) It is
`
`2 Oberg & Eriksson published in a journal, Best Practice & Research Clinical Gas-
`
`troenterology, directed at gastroenterologists, not oncologists. (Ex. 1028; Ex.
`
`1119, Ratain Supp. Decl. ¶49.) The more relevant disclosures related to new can-
`
`cer therapies would be included in journals directed to oncologists, like Oncologia,
`
`where Oberg 2004 published. (Id.)
`
`7
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`precisely these “new therapies” that Dr. Oberg identified as “experimental
`
`therapies” in his 2004 algorithm: “Gleevec, Herceptin, Rapamycin.” (Ex. 1027,
`
`Fig. 1; Ex. 1119, Ratain Supp. Decl. ¶50.) A POSA facing the problem of
`
`“develop[ing] a new therapy” (Ex. 2041, Kulke Decl. ¶118 (emphasis added))
`
`would not be limited to those treatments already in clinical use but would identify
`
`new drugs, like rapamycin and others based on tumor biology included in Oberg
`
`2004. (Ex. 1119, Ratain Supp. Decl. ¶50.)
`
`Contrary to Novartis’s arguments, Oberg & Eriksson does not teach away
`
`from using rapamycin in treating NETs. “The mere disclosure of alternative
`
`designs does not teach away,” and Oberg & Eriksson does not criticize, discredit,
`
`or discourage a POSA from “following the path set out in” Oberg 2004. In re
`
`Mouttet, 686 F.3d 1322, 1333-34 (Fed. Cir. 2012) (citations omitted). The path to
`
`the claimed method was clearly laid out in Oberg 2004, placing it well within the
`
`reach of a POSA.
`
`D. The ’224 Patent and Prior Art Group Rapamycin, Everolimus,
`and Temsirolimus as a Class of Anticancer Agents
`
`Novartis tries to manufacture a patentable distinction between everolimus
`
`and other rapalogs, but the prior art and admissions in the ’224 patent foreclose
`
`that argument. Smith & Nephew, Inc. v. Rea, 721 F.3d 1371, 1380 n.6 (Fed. Cir.
`
`2013) (stating that “[e]xpert opinions that are contrary to admissions in the
`
`8
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`specification do not create a factual issue”); PharmaStem Therapeutics, Inc. v.
`
`ViaCell, Inc., 491 F.3d 1342, 1361-62 (Fed. Cir. 2007) (disregarding testimony
`
`regarding state of the art that “is contrary to the representation in the
`
`specification”).
`
`The prior art consistently referred
`
`to rapamycin, everolimus, and
`
`temsirolimus as a class of new anticancer agents based on their mechanism of
`
`inhibiting mTOR. (Pet. 22-25, 28; Ex. 1003, Ratain Decl. ¶¶75-78, 86-87, 90-92;
`
`Ex. 1009, Dancey; Ex. 1018, Huang 2003; Ex. 1033, Rao; Ex. 1039, Vignot.) The
`
`pharmacological function, including antitumor activity, of mTOR inhibitors was
`
`expected to be similar to that of rapamycin. (Pet. 27-28; Ex. 1119, Ratain Supp.
`
`Decl. ¶58.) Indeed, Novartis recognized that rapamycin and its derivatives were a
`
`“class” in multiple prior art patents. (E.g., Ex. 1123, ’243 patent 1:50-55
`
`(“Rapamycin and its structurally similar analogs and derivatives are termed
`
`collectively as ‘compounds of the rapamycin class’” that “have also been shown to
`
`have antitumor . . . activity”); Ex. 1124, ’970 patent 36:51-52 (describing “mTOR
`
`inhibitor” as “e.g., rapamycin, [everolimus]”).)
`
`The ’224 patent admits that rapamycin, everolimus, and temsirolimus were
`
`known mTOR inhibitors possessing similar activities. For example, the patent
`
`states “[t]he compound rapamycin and other mTOR inhibitors inhibit mTOR
`
`activity via a complex with its intracellular receptor FKBP12.” (Ex. 1001 1:9-11.)
`
`9
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`The patent further identifies that “mTOR inhibitors” “preferably” include
`
`everolimus and temsirolimus. (Id. 2:21-34; Ex. 1119, Ratain Supp. Decl. ¶59.)
`
`The patent characterizes the state of the prior art, identifying that “mTOR
`
`inhibitors, on the basis of observed activity, have been found to be useful e.g. as
`
`immunosuppressants . . . and have additionally potent antiproliferative properties
`
`which make them useful for cancer chemotherapy, particularly for the treatment of
`
`solid tumors, especially of advanced solid tumors.” (Ex. 1001 2:35-40.)
`
`Novartis’s, and Dr. Kulke’s, attempts now to distinguish among this class “cannot
`
`be reconciled with statements made by the inventors in the [] specification and
`
`with the prior art references themselves.” PharmaStem, 491 F.3d at 1361.
`
`Accordingly, Novartis’s attempt to distinguish everolimus from its known
`
`closely related mTOR inhibitors should be disregarded.
`
`E. A POSA Would Not Have Expected PNETs “After Failure of
`Cytotoxic Chemotherapy” to Be “More Resistant”
`To avoid the clear teaching in the prior art of mTOR inhibitors as a class for
`
`the treatment of PNETs, Novartis attempts to elevate the claim limitation “after
`
`failure of cytotoxic chemotherapy” as representing a category of PNETs that would
`
`be “more resistant” to treatment. (POR 13-15.) But the prior art and Dr. Kulke’s
`
`own contemporaneous practice contradict Novartis’s assertion. The prior art
`
`shows that a POSA would not have expected PNETs that previously failed
`
`10
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`cytotoxic chemotherapy to be “more resistant” to an mTOR inhibitor like
`
`everolimus. (Ex. 1119, Ratain Supp. Decl. ¶¶27-41.)
`
`In support of his opinion that previously-treated PNETs would be considered
`
`“more resistant” to treatment, Dr. Kulke relies on four papers, each of which
`
`merely reflect the well-known general rule that PNETs, like all solid tumors, were
`
`less likely to respond to a second-line cytotoxic chemotherapy after failure of
`
`cytotoxic chemotherapy. (Id. ¶¶28-31, 34-36.) These references do not state that
`
`PNETs would be resistant to different treatment, such as molecularly targeted
`
`therapies like mTOR inhibitors. Instead, these references would encourage a
`
`POSA to use an anticancer agent with a completely different mechanism of action,
`
`such as an mTOR inhibitor, to treat such tumors. (Id. ¶32.) Indeed, Dr. Kulke
`
`previously agreed, testifying that when a PNET becomes refractory to treatment,
`
`“you have to look for another type of therapy.” (Id.; Ex. 1095, Kulke Sutent
`
`Testimony 798:11-21 (emphasis added).) Dr. Kulke’s previous testimony was
`
`correct. And the prior art confirms that molecularly targeted therapies exhibited
`
`antitumor activity in tumors previously treated with cytotoxic chemotherapy. (Ex.
`
`1119, Ratain Supp. Decl. ¶33.)
`
`Compelling contemporaneous evidence demonstrates that prior treatment
`
`with cytotoxic chemotherapy would not have been expected to affect responses to
`
`mTOR inhibitors. (Id. ¶37.) As of November 2005, multiple clinical studies
`
`11
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`assessing new treatments for NETs, including PNETs, frequently allowed patients
`
`to enroll who had previously been treated with cytotoxic chemotherapy. (Id.)
`
`Such inclusion criteria indicates that the investigators did not believe that prior
`
`treatment would likely affect the results of their investigation. (Id. ¶¶37-38.)
`
`Novartis and its expert do not identify a single prior art study separately analyzing
`
`the results of NET patients who had received prior cytotoxic chemotherapy. (Id.
`
`¶39.) A POSA would have expected any such difference to be reported. (Id.) In
`
`fact, the RADIANT-3 study reports treatment of these tumors “are comparable,”
`
`but makes no assertion that such “comparable” results were surprising or
`
`unexpected. (Id. ¶41.)
`
`Finally, the ’224 patent disclosure itself undermines Novartis’s position
`
`because it “adds nothing beyond the teachings” of the prior art. Merck & Co. v.
`
`Teva Pharm. USA, Inc., 395 F.3d 1364, 1374 (Fed. Cir. 2005). In reversing the
`
`district court, the Merck panel noted that the specification of the asserted patent
`
`“set[] forth no human clinical or laboratory data showing the safety and tolerability
`
`of the treatment methods claimed by the patent.” Id. Because “the claimed
`
`invention add[ed] nothing beyond the teachings” of the prior art, “the district court
`
`clearly erred in finding any difference between the claimed invention and the [prior
`
`art] on this point.” Id. Here, the ’224 patent provides no data regarding the
`
`efficacy of everolimus in PNETs, much less any data regarding PNETs previously
`
`12
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`failing cytotoxic chemotherapy, and therefore Novartis cannot point to any
`
`difference between the patent and the prior art on this point. (Ex. 1001; see Ex.
`
`1070, Kulke Dep. 114:16-115:8.) “[J]ust as a skilled artisan would be able to
`
`practice the invention claimed in the [] patent despite its lack of explicit” data
`
`regarding previously-treated tumors, “the artisan would have a reasonable
`
`expectation of success for adapting” Oberg 2004 or Duran “for the same use.”
`
`Alcon Research, Ltd. v. Apotex Inc., 687 F.3d 1362, 1369 (Fed. Cir. 2012).
`
`F. A POSA Would Have Reasonably Expected that Everolimus
`Would Be Therapeutically Effective in Advanced PNETs
`
`Next, in attacking a POSA’s expectation of success, Novartis would require
`
`a guarantee, but the law only requires that the expectation of success be
`
`reasonable. In re O’Farrell, 853 F.2d 894, 903 (Fed. Cir. 1988). “[O]bviousness
`
`cannot be avoided simply by a showing of some degree of unpredictability in the
`
`art.” Pfizer, Inc. v. Apotex, Inc., 480 F.3d 1348, 1364 (Fed. Cir. 2007). The
`
`evidence demonstrates that a POSA would have had a reasonable expectation of
`
`succeeding here.
`
`1.
`
`The role of mTOR inhibitors as anticancer agents, including
`in NETs, was not “uncertain”
`
`Oberg 2004 and Duran clearly identify the use of mTOR inhibitors to treat
`
`NETs, including PNETs. To avoid these clear teachings, Novartis argues that the
`
`etiology of PNETs was “poorly understood” and the role of mTOR inhibitors in
`
`13
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`treating cancer was “uncertain.” (POR 15-21.) But the question is not whether
`
`treating PNETs is generally unpredictable—the question is more narrowly whether
`
`there would be a reasonable expectation of success of using mTOR inhibitors to
`
`treat PNETs. See Allergan, Inc. v. Apotex Inc., 754 F.3d 952, 965 (Fed. Cir. 2014).
`
`In Allergan, the Federal Circuit reversed the district court to hold the
`
`claimed methods of growing hair obvious. Id. That “hair growth is generally an
`
`unpredictable endeavor” was not the relevant question. Id. The prior art disclosed
`
`using certain PGF analogs and thus “the question is more narrowly whether the
`
`success of using selective PGF analogs to treat hair loss would be reasonably
`
`unpredictable.” Id. Once a POSA was aware of the prior art teaching, “the general
`
`characteristics of the hair growth art ceased to be relevant.” Id.
`
`Here, as in Allergan, once a POSA was aware of the teachings of Oberg
`
`2004 and Duran, “the general characteristics” of treating PNETs ceased to be
`
`relevant. Id. Oberg 2004 taught rapamycin as an appropriate treatment for NETs,
`
`including PNETs. (Pet. 29-32.) Duran taught that temsirolimus had antitumor
`
`activity in NETs, including PNETs. (Id. 36-37.) Therefore, the correct question,
`
`at the time of the ’224 patent’s filing, is whether there was anything “unpredictable
`
`or mysterious” about an mTOR inhibitor treating PNETs. Given the clear teaching
`
`of these mTOR inhibitors as treatments for PNETs, the use of another mTOR
`
`inhibitor for the same purpose would not have been “mysterious.” (Ex. 1119,
`
`14
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`Ratain Supp. Decl. ¶¶64-71.)
`
`As in Allergan, Oberg 2004 and Duran did not make “a general exhortation
`
`covering thousands of possibilities.” 754 F.3d at 965. Instead, their “teaching[s]
`
`focused on specific classes of [compounds] with [known structures] and properties
`
`that guided persons of ordinary skill in the art to compounds with similar structures
`
`that would fall within the scope of the [’224] patent’s claims.” 754 F.3d at 965.
`
`“[T]here was little left to do but to confirm that the strategy suggested” by Oberg
`
`2004 and Duran “would work,” “provid[ing] a sound basis” that a POSA would
`
`reasonably expect everolimus would be effective. Genzyme Therapeutic Prods.
`
`L.P. v. Biomarin Pharm. Inc., 825 F.3d 1360, 1373 (Fed. Cir. 2016); (Pet. 40-45,
`
`49-50).
`
`Additionally, the role of mTOR inhibitors as anticancer agents in NETs was
`
`not “uncertain,” as Novartis and Dr. Kulke allege. (Ex. 1119, Ratain Supp. Decl.
`
`¶¶51-52, 60-63; POR 15-21; Ex. 2041 ¶¶66-99.) As of November 2005, mTOR
`
`had been
`
`identified “as an
`
`important
`
`target for anticancer
`
`therapeutics
`
`development” with reported antitumor activity for “all agents” in the class of
`
`mTOR inhibitors, including rapamycin, temsirolimus, and everolimus. (Ex. 2027
`
`Dancey 2005 313 & 323; Ex. 1119, Ratain Supp. Decl. ¶51.) Temsirolimus had
`
`advanced to multiple phase 3 trials and had been “granted . . . fast-track status” for
`
`renal cell carcinoma. (Ex. 1119, Ratain Supp. Decl. ¶51; Ex. 1115 at 363.) A
`
`15
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`POSA would have considered the advanced clinical development of mTOR
`
`inhibitors as firmly establishing their role as anticancer agents. (Ex. 1119, Ratain
`
`Supp. Decl. ¶51.)
`
`The clinical study in Duran demonstrates that mTOR inhibitors were
`
`identified as treatments for NETs. (Id. ¶52.) The National Cancer Institute
`
`(“NCI”) sponsored the Duran study, funding it after a competitive review by
`
`several committees that analyzed the supporting data and found the study
`
`warranted government funding. (Id.) Duran explained that a basis for the study
`
`was the connection between NETs and PTEN defects. (Id. ¶¶62, 64.) Thus,
`
`irrespective of other possible signaling pathways that may play a role in PNETs, a
`
`POSA would have been aware that sufficient evidence warranted the NCI granting
`
`scarce government resources to study treating NETs with mTOR inhibitors. (Id.
`
`¶¶52, 62-64.)
`
`A POSA thus would have been aware of this connection between NETs and
`
`mTOR because it had been suggested in the prior art, including Duran and Oberg
`
`2004. (Pet. 36-37, 40-42, 49-50; Ex. 1119, Ratain Supp. Decl. ¶¶60-64.) The role
`
`of mTOR inhibitors as anticancer agents, including in NETs, was anything but
`
`“uncertain.”
`
`16
`
`

`

`Case IPR2016-01479
`U.S. Patent No. 9,006,224
`
`
`2.
`
`A POSA would consider preclinical models for somatostatin
`receptor-positive tumors as models for NETs
`
`Trying to avoid the clear teaching of Boulay 2004 of everolimus’s antitumor
`
`activity in a model for NETs, Novartis and Dr. Kulke incorrectly contend that the
`
`CA20948 cell line allegedly is not such a model. (POR 30-38; Ex. 2041 ¶¶131-
`
`152.) Dr. Kulke bases this opinion chiefly by stating that CA20948 derives from
`
`(1) rats, (2) an adenocarcinoma of the exocrine pancreas, and (3) an azaserine-
`
`induced tumor. (Ex. 2041 ¶¶131-32, 134-35; Ex. 1070, Kulke Dep. 62:21-63:14;
`
`Ex. 1119, Ratain Supp. Decl. ¶44.)
`
`Dr. Kulke’s opinion here, however, is contradicted by his own prior
`
`publication. Specifically, Dr. Kulke co-authored a paper characterizing a tumor
`
`cell line (AR42J) derived from a rat azaserine-induced adenocarcinoma of the
`
`exocrine pancreas as a “Neuroendocrine Tumor Model.” (Ex. 1119, Ratain Supp.
`
`Decl. ¶44.) Just as Dr. Kulke identified AR42J as a NET model, a POSA would
`
`consider CA20948 a NET model. (Id.)
`
`Dr. Kulke admits that CA20948 is a model for somatostatin receptor-
`
`positive tumors, which include NETs. (Id. ¶¶42-43.) A POSA would have been
`
`aware that this cell line had been widely used for targeted therapies for NETs and
`
`would have associated activity in the cell line with activity in NETs. (Id