`
`2 4
`
`䡠 N U M B E R 1 6
`
`䡠
`
`J U N E
`
`1
`
`2 0 0 6
`
`JOURNAL OF CLINICAL ONCOLOGY
`
`O R I G I N A L R E P O R T
`
`From the University of Chicago,
`Chicago, IL; Royal Marsden Hospital,
`Surrey, United Kingdom; Abramson
`Cancer Center, University of Pennsylva-
`nia, Philadelphia, PA; The University of
`Texas M.D. Anderson Cancer Center,
`Houston; Cancer Therapy and Research
`Center, San Antonio, TX; Bayer Pharma-
`ceuticals Corporation, West Haven, CT.
`
`Submitted August 8, 2005; accepted
`December 1, 2005.
`
`Supported by Bayer Pharmaceuticals
`Corporation and Onyx Pharmaceuticals.
`
`Authors’ disclosures of potential con-
`flicts of interest and author contribu-
`tions are found at the end of this
`article.
`
`Address reprint requests to Mark J.
`Ratain, MD, University of Chicago,
`5841 S Maryland Ave, MC2115,
`Chicago, IL 60637; e-mail: mratain@
`medicine.bsd.uchicago.edu.
`
`© 2006 by American Society of Clinical
`Oncology
`
`0732-183X/06/2416-2505/$20.00
`
`DOI: 10.1200/JCO.2005.03.6723
`
`Phase II Placebo-Controlled Randomized Discontinuation
`Trial of Sorafenib in Patients With Metastatic Renal
`Cell Carcinoma
`Mark J. Ratain, Tim Eisen, Walter M. Stadler, Keith T. Flaherty, Stan B. Kaye, Gary L. Rosner, Martin Gore,
`Apurva A. Desai, Amita Patnaik, Henry Q. Xiong, Eric Rowinsky, James L. Abbruzzese, Chenghua Xia,
`Ronit Simantov, Brian Schwartz, and Peter J. O’Dwyer
`
`A
`
`B
`
`S
`
`T
`
`R
`
`A
`
`C
`
`T
`
`Purpose
`This phase II randomized discontinuation trial evaluated the effects of sorafenib (BAY 43-9006), an
`oral multikinase inhibitor targeting the tumor and vasculature, on tumor growth in patients with
`metastatic renal cell carcinoma.
`Patients and Methods
`Patients initially received oral sorafenib 400 mg twice daily during the initial run-in period. After 12
`weeks, patients with changes in bidimensional tumor measurements that were less than 25%
`from baseline were randomly assigned to sorafenib or placebo for an additional 12 weeks; patients
`with ⱖ 25% tumor shrinkage continued open-label sorafenib; patients with ⱖ 25% tumor growth
`discontinued treatment. The primary end point was the percentage of randomly assigned patients
`remaining progression free at 24 weeks after the initiation of sorafenib.
`Results
`Of 202 patients treated during the run-in period, 73 patients had tumor shrinkage of ⱖ 25%.
`Sixty-five patients with stable disease at 12 weeks were randomly assigned to sorafenib (n ⫽ 32)
`or placebo (n ⫽ 33). At 24 weeks, 50% of the sorafenib-treated patients were progression free
`versus 18% of the placebo-treated patients (P ⫽ .0077). Median progression-free survival (PFS)
`from randomization was significantly longer with sorafenib (24 weeks) than placebo (6 weeks;
`P ⫽ .0087). Median overall PFS was 29 weeks for the entire renal cell carcinoma population (n ⫽
`202). Sorafenib was readministered in 28 patients whose disease progressed on placebo; these
`patients continued on sorafenib until further progression, for a median of 24 weeks. Common
`adverse events were skin rash/desquamation, hand-foot skin reaction, and fatigue; 9% of patients
`discontinued therapy, and no patients died from toxicity.
`Conclusion
`Sorafenib has significant disease-stabilizing activity in metastatic renal cell carcinoma and is
`tolerable with chronic daily therapy.
`
`J Clin Oncol 24:2505-2512. © 2006 by American Society of Clinical Oncology
`
`INTRODUCTION
`
`Sorafenib (BAY 43-9006) is an oral kinase inhibitor
`targeting both tumor cells and the tumor vascula-
`ture. It was originally developed as an inhibitor of
`Raf-1, a member of the Raf/MEK/ERK signaling
`pathway.1,2 Sorafenib was subsequently found to
`have activity against B-Raf, vascular endothelial
`growth factor receptor–2, platelet-derived growth
`factor receptor, Fms-like tyrosine kinase-3 (Flt-3),
`and stem-cell growth factor (c-KIT).3 In phase I
`studies investigating various oral dosing schedules,
`sorafenib was generally well tolerated; the recom-
`mended dose for future trials was 400 mg bid con-
`
`tinuously. Dose-limiting toxicities at continuous
`doses higher than 400 mg bid were diarrhea, fatigue,
`and skin toxicity.4-7
`Preclinical studies in xenograft models (colon,
`breast, lung) showed that the primary effect of sor-
`afenib was inhibition of tumor growth rather than
`tumor shrinkage.3 These data suggested that, unlike
`cytotoxic agents, the primary clinical benefit of
`agents such as sorafenib may be disease stabilization.
`Therefore, classical oncology paradigms for phase II
`clinical evaluation (eg, single-arm noncontrolled
`studies using partial or complete response rate as
`the primary end point) would not adequately
`detect the activity of sorafenib.8 As duration of
`
`2505
`
`Downloaded from ascopubs.org by 38.94.143.76 on April 26, 2017 from 038.094.143.076
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2103
`Par v. Novartis, IPR 2016-01479
`Page 1 of 8
`
`
`
`Ratain et al
`
`disease stabilization is affected by the natural history of the disease
`and the effect of any administered agent, drug effect is best mea-
`sured through use of a placebo control, ideally with minimization
`of patient exposure to placebo.
`The randomized discontinuation (or withdrawal) trial (RDT)
`design, first proposed in 1975, attempts to assess the clinical activity of
`a drug while minimizing the use of placebo.9 Since then, this design
`has been used in many therapeutic areas.10-15 This is an enrichment
`design, in which all patients receive study drug for an initial run-in
`period, followed by random assignment of potential responders to
`either the study drug or placebo.9,14 This design creates a controlled
`trial without upfront randomization, and decreases the heterogeneity
`of the randomly assigned population, resulting in increased statistical
`power with smaller patient numbers. This design was first imple-
`mented in oncology in a study of carboxyaminoimidazole for the
`treatment of metastatic renal cell carcinoma (RCC).16,17
`Our multicenter placebo-controlled RDT was performed to de-
`termine whether sorafenib inhibits tumor growth in patients with
`metastatic solid tumors who maintain stable disease after a 12-week
`run-in period. The original protocol focused on patients with meta-
`static colorectal carcinoma (CRC), based on the putative importance
`of Raf/MEK/ERK signaling in this tumor type.18,19 However, the
`broad eligibility criteria of the protocol also enabled enrollment of
`patients with other malignancies. Early signs of antitumor activity in
`patients with RCC and low numbers of patients with CRC achieving
`the criteria for randomization after the 12-week run-in period led to a
`refocus of this study toward patients with RCC, as we have reported here.
`
`PATIENTS AND METHODS
`
`Patients
`Patients with histologically or cytologically confirmed metastatic refrac-
`tory cancer for which no approved effective therapy exists were eligible for this
`study. Originally, the study focused on patients with CRC, but allowed enroll-
`ment of patients with other solid tumor types. During the course of the study,
`evidence of tumor regression in many patients with RCC led to a protocol
`amendment, which extended recruitment of patients with RCC and termi-
`nated enrollment of patients with CRC.
`Inclusion criteria included: patient age of at least 18 years; at least one
`measurable tumor; Eastern Cooperative Oncology Group (ECOG) perfor-
`mance status of 0 or 1; life expectancy of at least 12 weeks; and adequate bone
`marrow, liver, and renal function. Patients with other serious medical prob-
`lems or CNS involvement were excluded. There was no limit on the extent of
`prior therapy, except for the exclusion of patients with previous exposure to a
`Ras pathway inhibitor.
`
`Study Design
`This RDT was conducted at five centers. Enrollment began on Septem-
`ber 25, 2002. This report includes efficacy data up to December 31, 2004. The
`study was conducted in accordance with the Declaration of Helsinki and Good
`Clinical Practice guidelines.
`Sorafenib (Bayer Pharmaceuticals Corporation, West Haven, CT) was
`initially administered to all patients in a 12-week open-label run-in period
`using continuous oral dosing at 400 mg bid. Doses of sorafenib were delayed or
`reduced if clinically significant toxicities considered related to sorafenib oc-
`curred. After the 12-week run-in period, disease status was assessed based on
`change in bidimensional tumor measurements from baseline.20 Patients with
`ⱖ 25% tumor shrinkage continued to receive sorafenib until disease progres-
`sion or toxicity, in order to avoid concerns about the random assignment of
`these patients. Patients with progressive disease (ⱖ 25% tumor growth or other
`evidence of progression) discontinued treatment. Patients who had a change in
`
`tumor size of less than 25% were randomly assigned to either sorafenib (at current
`dose) or matching placebo in a double-blinded fashion, using centralized alloca-
`tion via a telephone randomization system. Patients who progressed at any time
`after randomization (progression was defined as a change in bidimensional tumor
`measurement from randomization of ⱖ 25% or clinically assessed progression)
`were unblinded. Patients whose disease progressed while on placebo were offered
`sorafenib, and patients on sorafenib discontinued treatment.
`
`Assessment of Efficacy
`The primary end point was the percentage of randomly assigned patients
`who remained progression free at 12 weeks following random assignment (24
`weeks after study entry).
`Secondary end points included progression-free survival (PFS) after
`random assignment (randomized subset only); overall PFS (from start of
`treatment); tumor response rate; and safety. Tumor response was assessed at
`12 weeks, and once every 6 weeks thereafter, in accordance with modified
`WHO guidelines for partial response (PR), stable disease (SD), and progressive
`disease (PD). Objective responses were confirmed at least 4 weeks after the
`original documentation. In order to verify investigator observations in an unbi-
`ased manner, independent assessment of radiologic scans was performed retro-
`spectively for 152 (75%) of 202 patients. Some scans were not available for
`independent assessment, as a radiology charter specifying parameters for indepen-
`dent review was developed after the last patient was accrued. These independent
`radiographic assessments were performed by RadPharm (Princeton, NJ).
`
`Assessment of Safety
`Safety was assessed for the entire treatment period (run-in plus random-
`ization). All patients who received at least one dose of the study drug and who
`had post-treatment data available were assessable for safety. Safety assessments
`were performed every 3 weeks during the run-in and randomization phases,
`and once every 4 weeks thereafter. Toxicities were graded according to the
`National Cancer Institute Common Toxicity Criteria (version 2.0), and their
`relationship to the study drug was recorded.
`
`Statistical Analysis
`Simulations for computing power and sample size assumed that tumor
`growth was exponential and that the distribution of tumor growth rates was
`log-normal. The mean growth rate in these simulations led to 43% of patients
`with SD and 57% of patients with PD after 12 weeks, assuming no treatment
`effect. With 50 patients randomly assigned to each group, the study had a
`power of 81% to detect a drug effect that corresponded to a reduction in the
`progression rate from 90% to 70%, 12 weeks after randomization. This simu-
`lation did not consider the possibility of tumor shrinkage.16
`For the primary efficacy end point, the two treatment groups (based on
`an intention to treat) were compared using a Cochran–Mantel-Haenszel test
`stratified by baseline ECOG score; 95% CIs were computed using binomial
`distribution. PFS after randomization was summarized by the Kaplan-Meier
`method, and was compared between treatment groups using a log-rank test.
`We estimated PFS attributable to sorafenib by piecing together information
`from the various treatment groups and treatment periods. All patients contrib-
`uted to the PFS estimate for the first 12 weeks of therapy. We combined the PFS
`estimate for the first 12 weeks with a similar estimate for all remaining weeks
`after the first 12 weeks, the latter assuming the patient was alive and progres-
`sion free at 12 weeks. We estimated PFS after 12 weeks as a weighted average of
`group-specific PFS for the two groups treated with sorafenib for more than 12
`weeks: the 79 patients who entered the open-label part of the trial and the 33
`patients randomly assigned to continue on sorafenib. When combining the
`group-specific PFS estimates, the weights corresponded to the fraction of patients
`continuing on open-label sorafenib at 12 weeks (79 of 144 patients) and the
`proportion of randomly assigned patients who were progression free at 12 weeks
`(65 of 144 patients).
`
`RESULTS
`
`Thisstudydesignpermittedenrollmentofpatientswithavarietyoftumor
`types;502patientswereenrolledontothestudy,501ofwhomreceivedthe
`
`2506
`
`JOURNAL OF CLINICAL ONCOLOGY
`
`Downloaded from ascopubs.org by 38.94.143.76 on April 26, 2017 from 038.094.143.076
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2103
`Par v. Novartis, IPR 2016-01479
`Page 2 of 8
`
`
`
`Phase II RDT of Sorafenib in Metastatic RCC
`
`study drug. Early indications of activity in patients with RCC caused us to
`refocus our study on this patient population, resulting in RCC being the
`most predominant tumor type (202 patients [40%]).
`The baseline demographics of these RCC patients are listed in
`Table 1. In the randomized phase, the distribution of men and women
`differed between the treatment groups. However, there were no sig-
`nificant differences between groups for this or any of the other mea-
`sured baseline characteristics.
`
`Response Assessment: Run-in Phase
`Response was assessed at the end of the 12-week run-in based on
`investigator-assessed bidimensional tumor measurements. Response
`assessment was unavailable for nine patients (4%), all of whom had
`discontinued treatment before week 12. This response assessment was
`used to determine patients’ subsequent course of therapy. A total of 73
`patients (36%) achieved tumor shrinkage ⱖ 25% compared with
`baseline, 69 patients (34%) had tumor measurements that remained
`
`Table 1. Baseline Characteristics for All Treated Patients (n ⫽ 202)
`Patients by Random Assignment
`
`All Patients
`(N ⫽ 202)
`
`Placebo Group
`(n ⫽ 33)
`
`Sorafenib Group
`(n ⫽ 32)
`
`Characteristic
`
`Sex
`Male
`Female
`Age, years
`Median
`Range
`ECOG PS
`0
`1
`TNM stage
`I
`II
`III
`IV
`Missing
`Histologic subtype
`Clear cell
`Papillary
`Other
`Missing
`MSKCC risk categoryⴱ
`Low
`Intermediate
`High
`Missing
`No. of organ sites of disease
`1
`2
`ⱖ 3
`Sites of disease†
`Lung
`Lymph node
`Kidney
`Liver
`Duration of disease
`No. of patients
`No. of years
`Range
`Prior therapy
`Systemic anticancer therapy
`IL-2 or interferon
`Non-diagnostic surgery
`Radiotherapy
`Nephrectomy
`
`No.
`
`149
`53
`
`110
`92
`
`21
`49
`49
`68
`15
`
`152
`15
`11
`24
`
`69
`121
`6
`6
`
`32
`77
`93
`
`154
`86
`70
`52
`
`170
`154
`202
`68
`179
`
`58
`23-83
`
`198
`2.6
`0-21.9
`
`%
`
`74
`26
`
`54
`46
`
`10
`24
`24
`34
`7
`
`75
`7
`5
`12
`
`34
`60
`3
`3
`
`16
`38
`46
`
`76
`43
`35
`26
`
`84
`76
`100
`34
`89
`
`No.
`
`21
`12
`
`18
`15
`
`3
`6
`8
`15
`1
`
`25
`3
`2
`3
`
`14
`15
`3
`1
`
`4
`15
`14
`
`23
`16
`15
`10
`
`29
`28
`33
`11
`29
`
`60
`23-74
`
`33
`2.8
`0-11.7
`
`%
`
`64
`36
`
`55
`45
`
`9
`18
`24
`45
`3
`
`76
`9
`6
`9
`
`42
`45
`9
`3
`
`12
`45
`42
`
`70
`48
`45
`30
`
`88
`85
`100
`33
`88
`
`No.
`
`26
`6
`
`18
`14
`
`2
`11
`9
`8
`2
`
`27
`0
`1
`4
`
`13
`18
`0
`1
`
`8
`7
`17
`
`28
`14
`12
`5
`
`29
`26
`32
`9
`29
`
`58
`32-76
`
`31
`3.3
`0-21.2
`
`%
`
`81
`19
`
`56
`44
`
`6
`34
`28
`25
`6
`
`84
`0
`3
`13
`
`41
`56
`0
`3
`
`25
`22
`53
`
`88
`44
`38
`16
`
`91
`81
`100
`28
`91
`
`Abbreviatons: ECOG PS, Eastern Cooperative Oncology Group performance status; MSKCC, Memorial Sloan-Kettering Cancer Center (New York, NY); IL, interleukin.
`ⴱMSKCC risk category was assessed using four of the five original risk factors29 as follows: low Karnofsky performance status (⬍ 80%); low serum hemoglobin (⬍ lower
`limit of normal); high corrected serum calcium (⬎ 10 mg/dL); and absence of prior nephrectomy. High lactate dehydrogenase was omitted as a risk factor for the present
`study because lactate dehydrogenase measurements were not collected prospectively for all patients, and a more recent publication excluded high lactate dehydrogenase
`as an independent risk factor for survival.30 Risk categories were defined as: high risk, ⱖ 3 risk factors; intermediate risk, 1-2 risk factors; low risk, no risk factors.
`†Target or non-target lesions for ⬎ 20% of all 202 patients.
`
`www.jco.org
`
`2507
`
`Downloaded from ascopubs.org by 38.94.143.76 on April 26, 2017 from 038.094.143.076
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2103
`Par v. Novartis, IPR 2016-01479
`Page 3 of 8
`
`
`
`Ratain et al
`
`within 25% of baseline levels, and 51 patients (25%) showed either
`tumor growth ⱖ 25% or other evidence of progression at or before
`week 12 (Fig 1).
`Eight patients (4%) had independently confirmed PRs by mod-
`ified WHO criteria at 12 weeks, all of these patients continued on
`open-label treatment. Investigator-assessed PR rate by modified
`WHO criteria was 11%. Of the 15 patients treated with papillary
`cancer, investigator assessment of best response (using WHO criteria)
`showed two PRs at 12 weeks, with an additional three patients having
`tumor shrinkage of 25% to 49%.
`Patient Disposition
`The 12-week run-in was completed by 187 patients (93%). Of the
`15 patients who discontinued treatment before the 12-week assess-
`ment, the majority (12 patients) did so because of adverse events; one
`patient withdrew consent, one patient was lost to follow-up, and one
`patient died (as a result of pneumonia and metastatic disease, unre-
`lated to the study drug).
`Of the 69 patients identified at 12 weeks with tumor growth or
`tumor shrinkage of less than 25% who were eligible for entry onto
`the randomized phase, two patients continued on open-label sor-
`afenib (investigator protocol violation), and three patients withdrew
`(one patient each due to adverse events, to pursue other treatment
`options, and for clinical progression before random assignment). One
`patient who met the study criteria for PD at week 12 was randomly
`assigned instead of discontinuing treatment. Therefore, a total of 65
`patients were randomly assigned to receive sorafenib (32 patients) or
`placebo (33 patients). Seventy-three patients with tumor shrinkage of
`at least 25% at the 12-week assessment entered into the open-label part
`of the trial, plus six additional patients who continued sorafenib, either
`at the discretion of the investigator or after being granted a waiver,
`despite having SD (n ⫽ 3) or PD (n ⫽ 2), or not receiving treatment
`for the entire run-in (n ⫽ 1). Therefore, a total of 79 patients contin-
`ued open-label sorafenib. Forty-three patients, who completed the
`12-week run-in, discontinued treatment at a later time point; 40 pa-
`tients because of PD, and three patients who had SD (and withdrew
`from the study).
`
`Fig 2. Kaplan-Meier plot of investigator-assessed progression-free survival from
`week 12 randomization for patients randomized to placebo (n ⫽ 33) or to
`sorafenib (n ⫽ 32).
`
`Antitumor Activity
`Randomized phase. At 12 weeks postrandomization (24 weeks
`from study entry), 50% of patients (16 of 32 patients) receiving sor-
`afenib were progression free, compared with only 18%of patients (six
`of 33) receiving placebo (P ⫽ .0077). Median PFS from 12-week
`randomization was also statistically significantly longer in the sor-
`afenib group (24 weeks) compared with the placebo group (6 weeks;
`P ⫽ .0087; Fig 2).
`Sorafenib treatment was restarted in 28 patients whose disease
`progressed on placebo after a median time from randomization of 7
`weeks. The median time from restarting sorafenib to the end of treat-
`ment in these patients was 24 weeks, suggesting restabilization of PD.
`Entire treatment period. A secondary objective of this study was
`to estimate overall PFS for all treated patients. The 79 patients who
`continued on open-label sorafenib after 12 weeks had a median PFS
`from baseline of 40 weeks. In patients who achieved tumor shrink-
`age of at least 25% at 12 weeks (n ⫽ 73), PFS was not appreciably
`different in those patients who had tumor shrinkage of at least 25%
`to less than 50% (38 weeks; n ⫽ 45) with those patients who had
`tumor shrinkage of at least 50% (47 weeks; n ⫽ 28). This suggests
`that patients with minor tumor shrinkage may have the same
`
`Fig 1. Changes from baseline in investigator-assessed, bidimensional radio-
`graphic measurements at 12 weeks for patients with renal cell carcinoma. These
`measurements were unconfirmed, and therefore do not represent confirmed
`responses according to modified WHO criteria. Mean change at 12 weeks was
`⫺18% (standard deviation, 33%).
`
`Fig 3. Kaplan-Meier plot of estimated overall progression-free survival for all
`treated patients (n ⫽ 202) from day 1 of study drug dosing (excluding placebo-
`treated patient data). See Patients and Methods for details on the calculations.
`
`2508
`
`JOURNAL OF CLINICAL ONCOLOGY
`
`Downloaded from ascopubs.org by 38.94.143.76 on April 26, 2017 from 038.094.143.076
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2103
`Par v. Novartis, IPR 2016-01479
`Page 4 of 8
`
`
`
`Phase II RDT of Sorafenib in Metastatic RCC
`
`benefit as those with classic responses. For the entire population,
`median overall PFS was estimated (as described in Patients and
`Methods) to be 29 weeks (Fig 3).
`Safety
`The most common treatment-emergent adverse events were fa-
`tigue (73% of patients), rash/desquamation (66%), hand-foot skin
`
`reaction (62%), pain (other; 58%), and diarrhea (58%; Table 2). The
`majority of these events were grade 1 or 2 in severity, although nine
`patients discontinued drug because of toxicity. The most common
`grade 3/4 adverse event was hypertension, which was observed in 31%
`of patients. Antihypertensive therapy with a variety of agents was
`initiated in 46% of patients. No patients died from toxicity.
`
`Table 2. Incidence of Adverse Events Reported for at Least 10% of Patients in the Total Safety Population (N ⫽ 202)
`Any Grade
`Grade 3
`
`Grade 4
`
`Adverse Events
`
`Any adverse event
`Allergy/immunology
`Blood/bone marrow
`Hemoglobin
`Cardiovascular (general)
`Edema
`Hypertension
`Dermatology/skin
`Alopecia
`Dry skin
`Flushing
`Hand-foot skin reaction
`Dermatology/skin, other
`Rash/desquamation
`Constitutional symptoms
`Fever (in the absence of neutropenia)
`Fatigue (lethargy, malaise, asthenia)
`Weight loss
`Constitutional symptoms, other
`Gastrointestinal
`Anorexia
`Constipation
`Diarrhea, patients without colostomy
`Nausea
`Gastrointestinal, other
`Stomatitis/pharyngitis (oral/pharyngeal)
`Vomiting
`Renal/genitourinary
`Creatinine
`Hemorrhage
`Hepatic
`ALT
`AST
`Infection/febrile neutropenia
`Infection without neutropenia
`Musculoskeletal
`Metabolic/laboratory
`Hyperglycemia
`Hyperuricemia
`Hypophosphatemia
`Neurology
`Neuropathy, sensory
`Pain
`Abdominal pain or cramping
`Headache
`Arthralgia (joint pain)
`Myalgia (muscle pain)
`Pain, other
`Pulmonary
`Cough
`Pulmonary, other
`Dyspnea (shortness of breath)
`
`www.jco.org
`
`No. of
`Patients
`
`202
`21
`63
`54
`114
`30
`86
`187
`107
`47
`32
`125
`87
`134
`181
`24
`147
`66
`45
`192
`95
`65
`117
`61
`58
`70
`48
`50
`29
`45
`59
`22
`23
`75
`73
`29
`84
`34
`26
`31
`97
`40
`158
`39
`38
`25
`22
`117
`127
`57
`36
`77
`
`%
`
`100
`10
`31
`27
`56
`15
`43
`93
`53
`23
`16
`62
`43
`66
`90
`12
`73
`33
`22
`95
`47
`32
`58
`30
`29
`35
`24
`25
`14
`22
`29
`11
`11
`37
`36
`14
`42
`17
`13
`15
`48
`20
`78
`19
`19
`12
`11
`58
`63
`28
`18
`38
`
`No. of
`Patients
`
`108
`0
`13
`11
`69
`0
`62
`34
`0
`0
`0
`27
`0
`5
`17
`0
`12
`5
`0
`26
`6
`0
`8
`0
`6
`0
`0
`0
`0
`8
`10
`0
`0
`10
`10
`0
`25
`5
`0
`14
`8
`0
`22
`0
`0
`0
`0
`13
`17
`0
`6
`15
`
`%
`
`53
`0
`6
`5
`34
`0
`31
`17
`0
`0
`0
`13
`0
`2
`8
`0
`6
`2
`0
`13
`3
`0
`4
`0
`3
`0
`0
`0
`0
`4
`5
`0
`0
`5
`5
`0
`12
`2
`0
`7
`4
`0
`11
`0
`0
`0
`0
`6
`8
`0
`3
`7
`
`No. of
`Patients
`
`25
`0
`3
`3
`2
`0
`0
`0
`0
`0
`0
`0
`0
`0
`1
`0
`1
`0
`0
`2
`0
`0
`0
`0
`1
`0
`0
`0
`0
`0
`0
`0
`0
`0
`0
`0
`10
`1
`0
`0
`4
`0
`3
`0
`0
`0
`0
`2
`4
`0
`1
`3
`
`%
`
`12
`0
`1
`1
`1
`0
`0
`0
`0
`0
`0
`0
`0
`0
`⬍ 1
`0
`⬍ 1
`0
`0
`1
`0
`0
`0
`0
`⬍ 1
`0
`0
`0
`0
`0
`0
`0
`0
`0
`0
`0
`5
`⬍ 1
`0
`0
`2
`0
`1
`0
`0
`0
`0
`⬍ 1
`2
`0
`⬍ 1
`1
`
`2509
`
`Downloaded from ascopubs.org by 38.94.143.76 on April 26, 2017 from 038.094.143.076
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2103
`Par v. Novartis, IPR 2016-01479
`Page 5 of 8
`
`
`
`Ratain et al
`
`DISCUSSION
`
`The results of this placebo-controlled phase II study clearly demon-
`strate that sorafenib has significant activity in metastatic RCC, based
`on the marked difference in progression rate between patients ran-
`domly assigned to sorafenib versus placebo. Additional evidence for
`antitumor activity is provided by the restabilization of PD in patients
`whose disease had progressed on placebo and were switched to sor-
`afenib. As the median duration of sorafenib treatment in these patients
`after cross over was comparable to the median PFS for patients ran-
`domly assigned to placebo, this suggests that patients were not disad-
`vantaged from a brief period of placebo treatment, providing
`additional ethical support for this design.
`It is notable that if this study had been a single-arm study, the
`conclusion may have been that sorafenib was inactive in metastatic
`RCC, as a 4% PR rate is generally considered to be indicative of an
`inactive agent. Sorafenib has now been confirmed to have significant
`benefit in a large phase III trial, demonstrating a 100% prolongation of
`PFS, and a 39% prolongation of overall survival.21 Of note, in this
`phase III study, the confirmed response rate was also less than 5%.
`The RDT design also enables the assessment of tumor response in
`an uncontrolled setting by means of change in tumor measurements
`at the end of the run-in period (12 weeks). The majority of patients
`achieved some degree of tumor shrinkage at this time point, and
`tumor shrinkage of at least 25% (unconfirmed PR/minor response)
`was achieved in 36% of patients.
`Elucidation of the different histologic subtypes of RCC indicates
`that there are distinct molecular mechanisms responsible for tumor
`growth, and that patients with different subtypes may have variable
`prognoses and may respond differently to treatment.22 However, in
`this study, the antitumor effect in papillary RCC seemed similar to that
`of the clear cell RCC population.
`Although the RDT design prevents a direct analysis of PFS
`for patients initially treated with sorafenib, PFS of these patients
`was estimated by eliminating the number of patients randomly as-
`signed to placebo and combining the estimates of conditional PFS to
`avoid bias. Using this approach, the median PFS at 29 weeks was
`longer than that observed in patients treated with bevacizumab (21
`weeks), or with placebo in the same study (11 weeks).23 The lack of a
`
`difference in median PFS between patients with 25% to 50% tumor
`shrinkage, with those with tumor shrinkage of more than 50% (ie,
`partial responders), further indicates the limited utility of standard
`response criteria for development of agents such as sorafenib.8
`Importantly, this study shows that sorafenib is generally well
`tolerated when administered long term in a population of patients
`with metastatic cancer. Mild-to-moderate skin toxicity was common,
`and was reversed with treatment interruption and/or dose reduction.
`In addition, hypertension was frequent, and often required therapy
`with oral medications. It is expected that long-term sorafenib use will
`require administering antihypertensive therapy in many treated patients.
`Given that the drug is administered continuously without a break, the
`titration of antihypertensive therapy can be performed during the initial
`weeks of therapy and does not require frequent adjustment.
`Sorafenib has demonstrated activity as a single agent in other
`solid tumors. In our study, tumor shrinkage was evident at 12 weeks in
`patients with soft tissue sarcoma and thyroid cancer.24 Furthermore,
`sorafenib also appears to be active in advanced hepatocellular carcino-
`ma.25 The favorable safety profile of sorafenib lends itself well to
`combination with chemotherapy, as evidenced by the results of com-
`bining it with carboplatin and paclitaxel in metastatic melanoma.26
`This study was designed to evaluate the merits of the RDT design,
`which, as originally stated by Amery and Dony,9 allows increased
`statistical power and reduced patient numbers.27 Enrollment onto this
`trial was rapid, with 202 RCC patients enrolled in 16 months at five
`study sites, reflecting patient acceptance of the study design. The large
`treatment difference between sorafenib and placebo was demon-
`strated with only 65 randomly assigned patients, 35% fewer than
`originally estimated, because the actual treatment effect was much
`greater than the hypothesized treatment effect. Not only does this
`study design allow detection of antitumor activity by standard re-
`sponse rate criteria, it enables assessment of disease stabilization,
`which would not be possible with a standard phase II study design.
`Therefore, because only a relatively small proportion of treated pa-
`tients were randomly assigned, this underestimates the true efficiency
`of the RDT design. It is likely that the RDT design can be made even
`more efficient with modifications, such as using the full range of
`changes in tumor burden after random assignment, rather than a
`binary arbitrary criterion of “progression” as the primary end point.28
`
`REFERENCES
`
`1. Wilhelm S, Chien DS: BAY 43-9006: Preclini-
`cal data. Curr Pharm Des 8:2255-2257, 2002
`2. Hilger RA, Scheulen ME, Strumberg D: The
`Ras-Raf-MEK-ERK pathway in the treatment of can-
`cer. Onkologie 25:511-518, 2002
`3. Wilhelm SM, Carter C, Tang L, et al: BAY
`43-9006 exhibits broad spectrum oral anti-tumor activ-
`ity and targets the Raf/MEK/ERK pathway and recep-
`tor tyrosine kinases involved in tumor progression and
`angiogenesis. Cancer Res 64:7099-7109, 2004
`4. Awada A, Hendlisz A, Gil T, et al: Phase I
`safety and pharmacokinetics of BAY 43-9006 admin-
`istered for 21 days on/7 days off in patients with
`advanced, refractory solid tumours. Br J Cancer
`92:1855-1861, 2005
`5. Clark JW, Eder JP, Ryan D, et al: The safety and
`pharmacokinetics of the multi-targeted tyrosine kinase
`inhibitor (including Raf kinase and VEGF kinase), BAY
`
`43-9006, in patients with advanced, refractory solid
`tumors. Clin Cancer Res 11:5472-5480, 2005
`6. Moore MJ, Hirte HW, Siu L, et al: Phase I
`study to determine the safety and pharmacokinetics
`of the novel Raf kinase and VEGFR inhibitor BAY
`43-9006, administered for 28 days on/7 days off in
`patients with advanced, refractory solid tumors. Ann
`Oncol 16:1688-1694, 2005
`7. Strumberg D, Richly H, Hilger RA, et al: Phase
`I clinical and pharmacokinetic study of the novel Raf
`kinase and vascular endothelial growth factor recep-
`tor inhibitor BAY 43-9006 in patients with advanced
`refractory solid tumors. J Clin Oncol 23:965-972,
`2005
`8. Ratain MJ, Eckhardt SG: Phase II studies of
`modern drugs directed against new targets: If you
`are fazed, too, then resist RECIST. J Clin Oncol
`22:4442-4445, 2004
`9. Amery W, Dony J: A clinical trial design avoid-
`ing undue placebo treatment. J Clin Pharmacol
`15:674-679, 1975
`
`10. Chiron C, Dulac O, Gram L: Vigabatrin with-
`drawal randomized study in children. Epilepsy Res
`25:209-215, 1996
`11. Cibere J, Kopec JA, Thorne A, et al: Random-
`ized, double-blind, placebo-controlled glucosamine
`discontinuation trial in knee osteoarthritis. Arthritis
`Rheum 51:738-745, 2004
`12. Gotzsche PC, Hansen M, Stoltenberg M, et al:
`Randomized, placebo controlled trial of withdrawal
`of slow-acting antirheumatic drugs and of observer
`bias in rheumatoid arthritis. Scand J Rheumatol
`25:194-199, 1996
`13. Holmes C, Wilkinson D, Dean C, et al: The
`efficacy of donepezil in the treatment of neuropsy-
`chiatric symptoms in Alzheimer disease. Neurology
`63:214-219, 2004
`14. Kopec JA, Abrahamowicz M, Esdaile JM:
`Randomized discontinuation trials: Utility and effi-
`ciency. J Clin Epidemiol 46:959-971, 1993
`15. Uretsky BF, Young JB, Shahidi FE, et al:
`Randomized study assessing the effect of digoxin
`
`2510
`
`JOURNAL OF CLINICAL ONCOLOGY
`
`Downloaded from ascopubs.org by 38.94.143.76 on April 26, 2017 from 038.094.143.076
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2103
`Par v. Novartis, IPR 2016-01479
`Page 6 of 8
`
`
`
`Phase II RDT of Sorafenib in Metastatic RCC
`
`in patients with mild to moderate
`withdrawal
`chronic congestive heart failure: Results of the
`PROVED trial—PROVED Investigative Group. J Am
`Coll Cardiol 22:955-962, 1993
`16. Rosner GL, Stadler W, Ratain MJ: Randomized
`discontinuation design: Application to cytostatic anti-
`neoplastic agents. J Clin Oncol 20:4478-4484, 2002
`17. Stadler WM, Rosner G, Small E, et al: Suc-
`cessful implementation of the randomized discon-
`tinuation trial design: An application to the study of
`the putative antiangiogenic agent carboxyaminoimi-
`dazole in renal cell carcinoma—CALGB 69901. J Clin
`Oncol 23:3726-3732, 2005
`18. Bos JL: Ras