UNITED STATES PATENT AND TRADEMARK OFFICE
`
`_______________________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`_______________________________
`
`Amerigen Pharmaceuticals Limited and Argentum Pharmaceuticals LLC
`
`Petitioners
`
`v.
`
`Janssen Oncology, Inc.
`Patent Owner
`
`____________________
`
`
`U.S. Patent No. 8,822,438 to Auerbach et al.
`Issue Date: September 2, 2014
`Title: Methods and Compositions for Treating Cancer
`
`____________________
`
`
`Inter Partes Review No. 2016-002861
`____________________
`
`
`PETITIONERS’ REPLY TO PATENT OWNER’S
`RESPONSE TO PETITION
`
`
`
`
`
`
`
`
`1 Case IPR2016-01317 has been joined with this proceeding.
`
`
`
`

`

`
`
`Table of Contents
`Introduction ................................................................................................... 1
` Claim Construction .................................................................................... 1
` The Claims are a Combination of Two Known Elements for a
`Predictable Result ................................................................................................ 2
` A POSA would have expected that AA to treat prostate cancer might
`require co-administration of a GC based on the disclosures in the prior art
`regarding the predicted impact of AA on the adrenal steroid synthesis pathway . 2
`1. A POSA would have expected that AA would have an impact on adrenal
`reserve ............................................................................................................. 3
`2. O’Donnell demonstrates that administration of AA to a patient causes
`diminished cortisol response ........................................................................... 5
` The abnormal cortisol response for all patients in Study C of O’Donnell
`would have been interpreted as clinical evidence of the need for GC replacement
`therapy when administering AA to treat mCRPC ............................................... 6
` A POSA’s motivation to administer glucocorticoid replacement therapy to
`patients with prostate cancer treated with AA would have been informed by
`knowledge of the underlying pathophysiology and clinical manifestations of
`mCRPC; the presence of co-morbidities and the likelihood of concurrent stress
`in patients with mCRPC ..................................................................................... 8
` The prior art taught use of a glucocorticoid to treat symptoms associated
`with cortisol deficiency and chronic elevated ACTH .......................................... 9
` The prior art taught that use of glucocorticoid replacement therapy in
`patients with prostate cancer was safe ............................................................... 10
` Alleged Secondary Considerations Evidence Does Not Support
`Nonobviousness .................................................................................................. 12
` There Are No Unexpected Results ........................................................... 12
`1.
`Patent Owner Does Not Compare the Closest Prior Art ........................ 13
`2.
`Patent Owner Draws Conclusions from Irrelevant Data and Makes
`Improper Comparisons .................................................................................. 14
`3.
`Patent Owner Ignores Its Own Data That Shows The Combination of AA
`and Prednisone is Not Better ......................................................................... 16
`4.
`Patent Owner Ignores the Data that Prednisone May Contribute to Drug
`Resistance to AA ........................................................................................... 19
`
`
`
`

`

` The Commercial Success Achieved By Zytiga is Not the Result of the
`Claimed Invention ............................................................................................ 20
`1.
`The ‘213 Blocking Patent Renders Zytiga’s Marketplace Success
`Minimally Probative of Non-Obviousness .................................................... 20
`2.
`Patent Owner Fails to Establish Any Causal Nexus To The ‘438 Claims
`
`21
`3. Market Performance Of Zytiga Heavily Derives From Factors Other
`Than The Subject Matter Of The ‘438 Patent Claims .................................... 22
`
`
`
`
`
`
`
`
`
`

`

`
`
`Introduction
`
`The claims of the ‘438 patent are directed to treating prostate cancer by
`
`administering therapeutically effective amounts of abiraterone acetate (“AA”), a 17
`
`α-hydroxylase/C17,20-lyase inhibitor ("CYP17 inhibitor"), in combination with
`
`prednisone, a glucocorticoid. The prior art taught use of AA as an effective anti-
`
`cancer agent which suppresses testosterone synthesis in prostate cancer patients.
`
`AMG Ex. 1002, ¶¶ 26, 45, 56, 58. The prior art also taught that the combination of
`
`ketoconazole and prednisone is safe and effective in treating human patients with
`
`hormone-refractory advanced prostate cancer. AMG Ex. 1004, Abstract, pp. 1177-
`
`1178, 1179. A POSA would have been motivated to combine AA and a
`
`glucocorticoid, like prednisone, in view of the express teachings of the prior art.
`
`And the POSA would have expected that such a combination would result in a safe
`
`and effective treatment of prostate cancer. None of the arguments presented by the
`
`Patent Owner changes this conclusion.
`
` Claim Construction
`
`In the Institution Decision (“ID”), the Board construed a number of terms
`
`from the claims of the ’438 patent. Of those construed terms “treat,” “treating,”
`
`and “treatment” were construed to mean “include the eradication, removal,
`
`modification, management or control of a tumor or primary, regional, or metastatic
`
`cancer cells or tissue and the minimization or delay of the spread of cancer.” ID,
`
`
`
`1
`
`

`

`Paper 14 at 5. The Patent Owner agrees with this construction. PO Response at 8.
`
`Subsequent to filing their Response, the Patent Owner sought and received
`
`permission to file the Markman Order from the co-pending District Court
`
`litigation. Because the District Court construction of these terms is narrower than
`
`the construction in the ID, the District Court construction is not the “broadest
`
`reasonable construction in light of the specification,” and is thus irrelevant to the
`
`current proceeding. 37 C.F.R. § 42.100(b); see In re Cuozzo Speed Techs., LLC,
`
`793 F.3d 1268, 1275–79 (Fed. Cir. 2015), cert. granted sub nom. Cuozzo Speed
`
`Techs. LLC v. Lee, 136 S. Ct. 890 (mem.) (2016). Petitioners therefore believe no
`
`change in the claim construction from the ID is warranted, especially in view of the
`
`agreement of the Patent Owner to the construction in the Institution Decision. SAS
`
`Institute, Inc. v. ComplementSoft, LLC, 825 F.3d 1341, 1351 (Fed. Cir. 2016).
`
` The Claims are a Combination of Two Known Elements for a
`Predictable Result
`
` A POSA would have expected that AA to treat prostate cancer
`might require co-administration of a GC based on the disclosures
`in the prior art regarding the predicted impact of AA on the
`adrenal steroid synthesis pathway
`
`As Patent Owner admits, AA and ketoconazole are both CYP 17 inhibitors.
`
`(See Preliminary Response at Figures 1, 2). The prior art, including Barrie and
`
`O’Donnell, disclose that (1) the CYP 17 enzyme has two separate activities in the
`
`adrenal androgen synthesis pathway, a 17α-hydroxylase and a C 17,20 lyase
`
`
`
`2
`
`

`

`activity, and (2) AA is a more selective and potent inhibitor of both CYP 17
`
`activities than ketoconazole. AMG 1003 at 2318; AMG 1005, col. 22, ll. 61-66;
`
`col. 24, ll. 61-62; AMG 1003 at 2324; AMG 1188 at 114:6-128:13]; cf. JSN 2040,
`
`41, 42.
`
`Both Barrie (AMG 1005, col. 10, ll. 47-56, claims 2, 16, 21) and O’Donnell
`
`teach the use of AA to treat prostate cancer in a human patient, with O’Donnell
`
`specifically teaching that AA effectively suppresses adrenal androgen synthesis in
`
`men with mCRPC: “[t]hese studies demonstrate. . . the potential utility” of
`
`abiraterone in “causing reductions in testosterone levels in both castrate and
`
`noncastrate men with prostate cancer.” AMG 1003 at 2324. O’Donnell further
`
`teaches that “[a]drenocortical suppression may necessitate concomitant
`
`administration of replacement glucocorticoid.” AMG 1003, Abstract.
`
`The adrenal steroid synthesis pathway was well characterized as of August
`
`2006 (see, e.g., AMG 1026, Figure 1; AMG 1023 at 1242-43), and a POSA would
`
`have predicted that, as a result of AA’s 17α-hydroxylase inhibition, administration
`
`of AA to treat prostate cancer would suppress cortisol and increase ACTH drive.
`
`AMG 1093 at 14-19; AMG 1188 at 114:6-128:13; 171:3-174:11.
`
`1.
`
`A POSA would have expected that AA would have an
`impact on adrenal reserve
`
`A POSA would have expected that cortisol deficiency from administration
`
`of AA to treat prostate cancer would diminish adrenal reserve and potentially cause
`
`
`
`3
`
`

`

`adrenal insufficiency (AI), particularly during times of physiological stress. AMG
`
`1003, p.2323 (“an impact on adrenal reserve was predictable”); AMG 1093 at 20-
`
`24.
`
`In addition, a POSA would have expected that AA suppression of steroid
`
`synthesis downstream of CYP 17 would increase adrenal steroids upstream of CYP
`
`17, including mineralocorticoids such as DOC and weak glucocorticoids such as
`
`corticosterone. JSN 2040 at 33 (it was “expected” that corticosterone would still
`
`be made following administration of abiraterone); id. at 45, 46 (a POSA would
`
`“understand” that administration of AA would not interfere with “continued
`
`production of corticosterone”; AMG 1188 at 126:10-129:4; AMG 1093 at 25.
`
`A POSA also would have known that elevated levels of mineralocorticoids
`
`could cause mineralocorticoid excess, a condition characterized by hypertension,
`
`hypokalemia and fluid retention. AMG 1002, ¶31; AMG 1023. Contrary to his
`
`written opinion here (JSN 2040, ¶28), a POSA would have considered, by analogy,
`
`the clinical presentation of patients with combined congenital CYP 17 deficiency
`
`for guidance with respect to the anticipated clinical symptoms and appropriate
`
`options for managing such symptoms in patients administered AA to treat mCRPC,
`
`as Dr. Auchus has opined outside of this proceeding. AMG 1085, p 508; AMG
`
`1188 at. 126:10-129:4; AMG 1023 at 1243-44; 1245 and Fig. 1; AMG 1002 at 31-
`
`32; AMG 1093 at 24, 25; Cf. Patent Owner Response at ¶26-28. As of August
`
`
`
`4
`
`

`

`2006, combined congenital CYP 17 deficiency was a well-characterized condition
`
`in which patients lack or have diminished CYP 17 α-hydroxylase and CYP 17,20
`
`lyase activity (AMG 1026; AMG 1027; AMG 1025).
`
`2. O’Donnell demonstrates that administration of AA to a
`patient causes diminished cortisol response
`
`O’Donnell expressly states that the administration of AA to treat a patient
`
`with prostate cancer may require co-administration of a glucocorticoid as
`
`replacement therapy based on abnormal cortisol responses in all patients in Study
`
`C (the only multi-dose study reported in O’Donnell). AMG 1003 at 2321 and 2323
`
`(“Although baseline cortisol levels remain normal, all patients treated at 500 and
`
`800 mg in the multiple dose study developed an abnormal response to a short
`
`Synacthen test by Day 11. Some impact on adrenal reserve was predictable from
`
`the steroid synthesis pathway.” (emphasis added)).
`
`Faced with this clear teaching in the prior art, Patent Owner attempts to
`
`argue that those clear statements do not actually mean what they say, and that a
`
`POSA would have disregarded these unequivocal teachings, performed their own
`
`analysis of the underlying data, and reached a contrary conclusion. See PO
`
`Response at 39-41. The evidence, however, does not support Patent Owner’s
`
`attempt to undo the teachings of the prior art. Instead, a POSA would have
`
`interpreted the abnormal cortisol response for all patients in Study C of O’Donnell as
`
`
`
`5
`
`

`

`clinical evidence of the need for GC replacement therapy when administering AA to
`
`treat mCRPC. AMG 1091 at 30; AMG 1093 at 22, 37-64.
`
`
`
`The abnormal cortisol response for all patients in Study C of
`O’Donnell would have been interpreted as clinical evidence of the
`need for GC replacement therapy when administering AA to treat
`mCRPC
`
`As an initial matter, it is unlikely that a POSA would have had the in-depth
`
`knowledge of endocrinology required to parse the ACTH stimulation test results
`
`reported (AMG 1093¶¶ 34-60). Moreover, Patent Owner has provided no rationale
`
`why a POSA would have questioned them, particularly in view of the fact that, as
`
`O’Donnell also states, “an impact on adrenal reserve” was “predictable” based on
`
`the selective mechanism of action of AA. AMG 1003 at 2323; AMG 1091 at 30;
`
`35-36; AMG 109320-24. There simply would have been no reason for a POSA
`
`to question results (1) presented in a peer-reviewed journal, (2) consistent with the
`
`hypothesis, (3) predictable based on AA’s mechanism of action way, and (4)
`
`confirming what a POSA already would have been concerned about regarding the
`
`likely suppression of cortisol production and increase in ACTH by AA when
`
`administered to treat prostate cancer in a patient. AMG 1188 at 166:4-168:19.
`
`Even assuming, arguendo, that a POSA had consulted with an expert in
`
`endocrinology in interpreting these results, an endocrinologist would have arrived
`
`at the identical conclusion that O’Donnell reaches: that the results for the ACTH
`
`stimulation test for all patients in Study C are unquestionably abnormal, and that
`
`
`
`6
`
`

`

`these abnormal results imply that AA could result in chronic cortisol deficiency for
`
`which glucocorticoid replacement therapy could be required. AMG 1093 at 22,
`
`68; AMG 1023 at 1245.
`
`None of the purported “clinical evidence” cited by Drs. Rettig and Auchus
`
`as teaching away from these conclusions would have persuaded the POSA to reach
`
`a different conclusion. AMG 1093 at 20-24, 65-70; AMG 1091 at 38-39. For
`
`example, a POSA would have understood that, because Studies A and B in
`
`O’Donnell involved administration of a single dose of AA in sub-therapeutic
`
`amounts ranging from 10 mg to 500 mg AA (in Study A) or 200 mg or 500 mg (in
`
`Study B), no inferences could or should be drawn from these studies regarding the
`
`safety of abiraterone to treat prostate cancer. AMG1093 at 65-70; AMG 1091 at
`
`31; cf. JSN 2040 at 23-24, 31, 33, 58. A POSA also would have understood that
`
`clinical manifestations of cortisol deficiency were unlikely to occur over the short
`
`course of the studies in O’Donnell, including Study C, and could be masked by
`
`other conditions. AMG1093 at 23-25, 61, 68 69; AMG 1091 at 38, 39; cf. JSN
`
`2040 at 31, 33.
`
`Similarly, that O’Donnell does not discuss mineralocorticoid excess would
`
`not have changed the POSA’s expectation that long-term administration of AA
`
`could lead to the development of clinically relevant mineralocorticoid excess in
`
`such patients. AMG 1093 at 24-26, 64-65; Cf. JSN 2040 at 26, 58. Instead, the
`
`
`
`7
`
`

`

`POSA would have understood that the absence of clinical symptoms of
`
`mineralocorticoid excess in the short duration studies in O’Donnell did not mean
`
`that mineralocorticoid excess was not still a concern and that factors such as
`
`underlying conditions and sodium intake would be expected to influence the time
`
`of onset and severity of clinical and laboratory manifestations of mineralocorticoid
`
`excess in this context. AMG 1093, 70; cf. JSN 2040 at 26, 58.
`
` A POSA’s motivation to administer glucocorticoid replacement
`therapy to patients with prostate cancer treated with AA would
`have been informed by knowledge of the underlying
`pathophysiology and clinical manifestations of mCRPC; the
`presence of co-morbidities and the likelihood of concurrent stress
`in patients with mCRPC
`
`A POSA would have had all the more reason to proceed with caution in
`
`managing cortisol deficiency in patients with mCRPC because of the greater risk
`
`for adverse events from cortisol deficiency, including adrenal insufficiency and
`
`adrenal crisis, in this frail, older population. AMG 1093 at 20-27; AMG 1091 at
`
`35. A POSA would have understood that risks of adrenal insufficiency would be
`
`far greater among patients with prostate cancer as a result of the underlying
`
`pathology of the condition, their older age, limited mobility, and greater
`
`susceptibility to episodes of acute illness, (AMG 1093 at ¶¶20-24; AMG 1091 at
`
`35), even if Dr. Auchus did not consider this. Cf. AMG 1188 at 104:25-136:2-7.
`
`In addition, the POSA would have understood that a diagnosis of adrenal
`
`insufficiency in patients with mCRPC could be particularly challenging and easy to
`
`
`
`8
`
`

`

`miss for a host of reasons including the expected overlap between clinical
`
`manifestations of adrenal insufficiency and the underlying cancer and potential
`
`masking of clinical manifestations of adrenal insufficiency by clinical
`
`manifestations of secondary mineralocorticoid excess. AMG 1093 at 68-69. In this
`
`clinical setting, a POSA would have understood that clinically relevant
`
`glucocorticoid deficiency could develop in patients treated with AA in the absence
`
`of clinical manifestations of adrenal insufficiency, such as hypotension, orthostatic
`
`hypotension, and tachycardia. AMG 1093 at 20-24, 64-65; Cf. JSN 2040 at 21, 31,
`
`33.
`
` The prior art taught use of a glucocorticoid to treat symptoms
`associated with cortisol deficiency and chronic elevated ACTH
`
`To address these concerns, the POSA would have referred to the example of
`
`ketoconazole, as O’Donnell does, AMG 1188 at 168:22-169:23, and been guided
`
`by the teachings in the art, including Gerber that ketoconazole, co-administered
`
`with prednisone, was safe and effective to treat prostate cancer. AMG 1002 at 33-
`
`36, 69, 70; AMG 1091 at 37, 39; AMG 1004 at 1179 (ketoconazole “as a single
`
`agent has limited use” but “in contrast,” a subgroup of mCRPC patients may
`
`“derive significant benefit” from the “combination of ketoconazole and
`
`physiological glucocorticoid replacement therapy”).
`
`
`
`9
`
`

`

`
`
`The prior art taught that use of glucocorticoid replacement
`therapy in patients with prostate cancer was safe
`
`Contrary to Dr. Rettig’s opinion, the prior art manifestly did not teach away
`
`from co-administering low-dose glucocorticoids with AA to treat patients with
`
`mCRPC. AMG 1093 at 28-30; AMG 1091 at 41. Indeed, the consensus in the art
`
`was unequivocally to the contrary. AMG 1093 at 28-30; AMG 1091 at 41. It is
`
`certainly Dr. Auchus’ reading of the prior art. For example, outside of this
`
`proceeding, Dr. Auchus has concluded that the low-dose prednisone with AA is
`
`not associated with the adverse safety profile sometimes seen when treating other
`
`malignancies and autoimmune disorders with therapeutic glucocorticoid doses.
`
`AMG 1086, Abstract, p. 1235(noting that it is the “higher doses [of glucocorticoid
`
`doses] that have been consistently associated with an adverse safety profile.”);
`
`AMG 1188 at 153:22-154:10.
`
`
`
`Moreover, concern with side effects from long-term administration of
`
`glucocorticoids would not been a factor for a POSA because of the short survival
`
`prospects for patients with mCRPC (less than two years) and the enhanced risk of
`
`immediate and potentially fatal consequences if glucocorticoid replacement
`
`therapy was not co- administered. AMG 1091 at 40; AMG 1093 at 23, 28, 29.
`
`Mineralocorticoid excess, predictable from AA’s mechanism of action,
`
`would have provided a further motivation to co-administer a glucocorticoid. The
`
`
`
`10
`
`

`

`POSA would have known that that low-dose glucocorticoid replacement therapy
`
`was the “cornerstone of therapeutic management of symptoms” resulting from
`
`cortisol deficiency and chronic elevated ACTH in patients with combined
`
`congenital CYP 17 deficiency. AMG 1026 at 115; AMG 1002 at 32; AMG 1093
`
`at 26.
`
`No POSA would seriously have considered any of the therapeutic options
`
`presented by Dr. Auchus as preferable to a glucocorticoid in the clinical context of
`
`mCRPC. AMG 1093 24-30; 32-35. A POSA would have known that use of a
`
`mineralocorticoid antagonist could be counterproductive and dangerous if the
`
`patient developed adrenal crisis requiring glucocorticoid replacement therapy.
`
`AMG 1093 at 26. A POSA would have known that spironolactone was reported
`
`to activate androgen receptors (AR). AMG 1093 at 26. Potassium canrenoate,
`
`never commercially available in the United States, AMG 1188 at 112:3-23, was a
`
`known metabolite of spironolactone, and would have been expected to have a
`
`similar risk of AR activation. AMG 1093 at 27.
`
`In August 2006, eplerenone was a new drug agent, so that its use in
`
`managing symptoms of AA in patients with mCRPC would have been exploratory.
`
`AMG 1093 at 27. In addition, in the context of a new drug agent like AA seeking
`
`regulatory approval, the use of eplerenone was inadvisable at least because
`
`eplerenone’s mechanism of action was not fully understood and because federal
`
`
`
`11
`
`

`

`regulators were likely to have limited familiarity with eplerenone2. AMG 1093 at
`
`17-19; 31-34.
`
`In addition, none of these strategies would address the underlying cortisol
`
`deficiency. AMG 1093 at 26-27; 31-35. Because a POSA would have understood
`
`that cortisol deficiency from AA to treat mCRPC could (and indeed does)3 cause
`
`adrenal insufficiency, the prudent option to manage the predictable impact on
`
`adrenal reserve would have been co-administration of glucocorticoid replacement
`
`therapy. AMG 1093 20-24, 26-27, 34-35
`
`In this context, prednisone would have been an obvious choice among
`
`glucocorticoids because of its common use in patients with prostate cancer. AMG
`
`1002 at 68, 79; AMG 1091 at 36; AMG1093 at 28.
`
` Alleged Secondary Considerations Evidence Does Not Support
`Nonobviousness
`
` There Are No Unexpected Results
`
`In the present case, it was known that both AA and prednisone could be used
`
`in the treatment of prostate cancer, and it was entirely expected that the
`
`combination of both could be used to treat the same condition. AMG1091 at 47.
`
`
`2 A POSA’s cautious approach would find validation in later studies of
`eplerenone showing that, like spironolactone, eplerenone also activates mutant AR.
`Richards 2012. In addition, in later phase II trials of AA, it has been demonstrated
`that eplerenone can cause hypotension following glucocorticoid replacement.
`AMG 1093 at 26; AMG 1170 at 1486.
`3 AMG 1018 at Section 5.2.
`
`
`
`12
`
`

`

`As evidence of “unexpected benefit in treating the cancer” Patent Owner cherry-
`
`picks data that is irrelevant to what PO must demonstrate. The appropriate
`
`analysis for unexpected results is the comparison of AA plus prednisone to AA
`
`alone, since the latter was explicitly disclosed in O’Donnell and/or Barrie.
`
`1.
`
`Patent Owner Does Not Compare the Closest Prior Art
`
`“To be particularly probative, evidence of unexpected results must establish
`
`that there is a difference between the results obtained and those of the closest prior
`
`art, and that the difference would not have been expected by one of ordinary skill
`
`in the art at the time of the invention.” Bristol-Myers Squibb v. Teva Pharm. USA,
`
`752 F.3d 967, 977 (Fed. Cir. 2014). In this case, the closest prior art is O’Donnell
`
`and/or Barrie, each of which teach that AA can be used to treat prostate cancer.
`
`Thus, the proper comparison to establish that the present claims achieved some
`
`unexpected result would be a comparison to the treatment of AA alone. AMG1091
`
`at 47. Instead, Patent Owner compares data from a study comparing (1) AA and
`
`dexamethasone to AA alone; and (2) AA and prednisone to prednisone alone.
`
`AMG 1091 at 48-49. Perhaps this is not surprising, since Dr. Rettig admitted he
`
`did not consider any prior art apart from that provided by counsel for Patent
`
`Owner. AMG 1189 at 16:2-10. For this reason alone, Dr. Rettig’s incomplete
`
`analysis should be disregarded.
`
`
`
`13
`
`

`

`2.
`
`Patent Owner Draws Conclusions from Irrelevant Data and
`Makes Improper Comparisons
`
`In addition to not comparing to the closest prior art, Patent Owner also
`
`makes improper comparisons to irrelevant data. As Dr. Ratain explains Patent
`
`Owner’s evidence of purported unexpected results is based on comparisons
`
`between (1) AA and dexamethasone vs. AA alone; and (2) AA and prednisone vs.
`
`prednisone alone. AMG 1091 at 48-49; 51-52. However, the present claims are
`
`specifically directed to the combination of AA and prednisone, not
`
`dexamethasone.4 Cf. JSN 2038 at 177 (incorrectly stating “[t]he Attard (2008)
`
`and Attard (2009) publications show that the use of [AA] and prednisone in
`
`combination results in increased efficacy”). While Patent Owner attempts to argue
`
`that the results to dexamethasone can be extrapolated to include all
`
`glucocorticoids, the prior art clearly demonstrates this is scientifically erroneous.
`
`AMG 1091 at 52.
`
`
`4 This also is not surprising since Patent Owner has a history of
`mischaracterizing data when it suits their position. For example, in arguing to
`obtain the ‘438 Patent, Patent Owner made false statements regarding the teachings
`of Tannock 1996 arguing that, based on no significant difference in OS between
`mitoxantrone and prednisone vs. prednisone the POSA would have expected “there
`to be no significant difference in survival between one cancer agent alone and that
`same cancer agent in combination with prednisone.” AMG 1012 at 6. What PO
`conveniently overlooked is that because neither mitoxantrone nor prednisone alone
`had ever been shown to increase OS, a POSA would have had no reason to expect
`that the two agents together would increase OS.
`
`
`
`14
`
`

`

`First, Dr. Rettig ignores the prior art teachings that different glucocorticoids
`
`can differ in their effect on prostate cancer (AMG 1091 at 41, 58; JSN 2024 cf.
`
`Rettig JSN 2038 132-134 .
`
`Even more significant is Dr. Rettig’s omission of the extensive prior art
`
`showing that dexamethasone was an effective monotherapy treatment for prostate
`
`cancer, resulting in durable PSA response in patients with mCRPC. AMG 1091 at
`
`43, 59; AMG1121; AMG1122; AMG1123; AMG1124; AMG1124, AMG1125.
`
`Because dexamethasone was a known effective treatment for prostate cancer, it
`
`would not have been unexpected that the combination of dexamethasone and AA
`
`could be more effective than AA alone. AMG 1091 at 60.
`
`In addition, Dr. Rettig ignored Patent Owner’s own clinical data showing
`
`that dexamethasone plus AA is superior to prednisone plus AA, (AMG 1091 at
`
`61; AMG11285; AMG 1187 6) and clinical evidence demonstrating that
`
`substituting dexamethasone for prednisone results in a durable PSA response in
`
`patients with mCRPC who had progressed on AA with prednisone. AMG 1091 at
`
`
`5 An inventor of the ‘438 patent (Dr. de Bono) participated in this trial,
`sponsored by Janssen, so there is no question PO was aware of this publication at
`the time of filing its Response.
`6 The very title of this study, NCT 01867710, summarizes Patent Owner’s
`own assessment of the rationale for co-administration of prednisone with AA:
`“Abiraterone With Different Steroid Regimens for Side Effect Related to
`Mineralcorticoid Excess Prevention in Prostate Cancer Prior to Chemotherapy”.
`AMG 1187.
`
`
`
`
`
`15
`
`

`

`62; AMG 1129; AMG 1130; AMG 1089. Also overlooked is the data
`
`demonstrating that dexamethasone alone is superior to prednisolone (which is
`
`converted to prednisone) alone in treating patients with mCRPC, as measured by
`
`decreases in serum PSA. AMG 1091 at 63; AMG 1132. Because the clinical
`
`evidence demonstrates that prednisone and dexamethasone are not interchangeable
`
`in treating prostate cancer, data on the combination of AA plus dexamethasone is
`
`irrelevant to support unexpected results of the claims, directly only to AA and
`
`prednisone. AMG 1091 at 64.
`
`3.
`
`Patent Owner Ignores Its Own Data That Shows The
`Combination of AA and Prednisone is Not Better
`
`Perhaps most egregiously, Patent Owner, in cherry-picking data for Dr.
`
`Rettig, ignored its own data which demonstrates that the combination of AA and
`
`prednisone is not only not better, but may actually be worse, than what one would
`
`expect by combining the two.
`
`Had Dr. Rettig reviewed all of the available clinical data on AA and
`
`prednisone, he would have found data on later and much larger studies than the
`
`study he cites to. AMG 1091 at 53. In particular, he would have found a 2013
`
`report by Ryan (AMG1111) and a 2014 report by Rathkopf (AMG1109). AMG
`
`1091 at 53-56. Each of which provides data on the COU-AA-302 Phase 3 study of
`
`AA plus prednisone vs. prednisone in patients without prior chemotherapy, which
`
`is the same regimen used in Ryan 2011 and analyzed by Dr. Rettig in Table 1 of
`
`
`
`16
`
`

`

`his Declaration. AMG 1091 at 53; AMG 1111; AMG1109 . That data reports the
`
`time to PSA progression was only 11.1 months, much shorter than the 16.3 months
`
`reported in Ryan 2011, and included in Table 1 of his declaration.”7 AMG 1091 at
`
`53.
`
`Furthermore, Dr. Rettig did not consider the time to PSA progression for
`
`prednisone alone including Patent Owner’s own data reported in Ryan 2013.
`
`While prednisone has not been proven to improve overall survival, it was well-
`
`known as of August 2006 that treatment with prednisone monotherapy can cause a
`
`durable PSA response in patients with prostate cancer. AMG 1091 at 54;
`
`AMG1112, AMG1114, AMG1113.
`
`Indeed, Patent Owner’s own clinical data demonstrate that prednisone
`
`monotherapy results in a decrease in PSA. AMG 1091 at 54. For example, Ryan
`
`2013 (AMG 1111) and Rathkopf 2014 (AMG 1109) report that 29% of patients
`
`treated with prednisone alone had a 50% reduction in PSA. AMG 1091 at 54;
`
`Ryan 2013. This data is consistent with the 22% PSA response (50% reduction in
`
`
`7 Dr. Rettig also improperly concludes that an increase in time to PSA
`progression is evidence of a “clinical benefit”. As Dr. Ratain explains, “neither a
`decrease in serum PSA levels nor time to PSA progression (regardless of how it is
`measured) is a clinical benefit, or a validated surrogate endpoint of any clinical
`benefit, resulting from treatment of a patient with mCRPC. In other words, PSA-
`based endpoints are useful in the context of early drug development, but are not
`sufficient to support findings of effectiveness required for FDA approval.” AMG
`1091 at 57.
`(Continued...)
`
`
`
`17
`
`

`

`PSA) for prednisone alone reported in the Phase 3 study of mitoxantrone plus
`
`prednisone versus prednisone by Tannock in 1996 (AMG 1006).8 Because Dr.
`
`Rettig did not consider the data for prednisone alone, he did not need to explain
`
`why an improvement in time to PSA progression from the combination of AA and
`
`prednisone that is less than the additive response for AA and prednisone alone
`
`AMF 1091 at 53-56) is unexpected. However, adding the data for time to PSA
`
`progression for prednisone alone as reported in Ryan 2013 (i.e., 5.5 months),
`
`instead of that for dexamethasone as reported in Attard 2009 (Cf. JSN 2038 Table
`
`1), and the median time to progression for AA alone of 7.5 months from Attard
`
`2009 results in a median time to progression of 13.0 months, almost two months
`
`longer that the reported time to PSA progression for AA and prednisone of 11.1
`
`months in Ryan 2013. AMG 1091at 53-55. Thus, the incremental effect on time
`
`to PSA progression of AA in the setting of prednisone is approximately 5.5 months
`
`(11.1 minus 5.6), almost two months less than the median time to progression for
`
`AA alone of 7.5 months reported in Attard 2009 . AMG 1091 at 55. Thus if
`
`anything, the effects of AA on time to PSA progression are diminished in
`
`combination with prednisone.
`
`
`8 The results reported by Tannock 1996 (AMG 1006) are the basis for FDA
`approval of mitoxantrone to treat patients with mCRPC, as Dr. Rettig confirmed in
`his deposition. AMG 1189 at 145:16-148:19. Nevertheless, Dr. Rettig, failed to
`consider this information when forming his opinion.
`
`
`
`18
`
`

`

`4.
`
`Patent Owner Ignores the Data that Prednisone May
`Contribute to Drug Resistance to AA
`
`While Dr. Rettig opines that the combination of AA and