`
`JOURNAL OF CLINICAL ONCOLOGY
`
`O R I G I N A L R E P O R T
`
`Author affiliations appear at the end of
`this article.
`
`Published online ahead of print at
`www.jco.org on October 14, 2013.
`
`P.A. was supported by an American
`Society of Hematology (ASH) Scholar
`Award and by an American Society of
`Clinical Oncology/Conquer Cancer Foun-
`dation Career Development Award.
`D.A.R. was supported by an ASH
`Scholar Award.
`
`P.A. and A.N. contributed equally to
`this work.
`
`Authors’ disclosures of potential con-
`flicts of interest and author contribu-
`tions are found at the end of this
`article.
`
`Clinical trial information: NCT00532259.
`
`Corresponding author: Leo I. Gordon,
`MD, Northwestern University Feinberg
`School of Medicine and the Robert H.
`Lurie Comprehensive Cancer Center,
`676 North St Clair Street, Suite 850,
`Chicago, IL 60611-3008; e-mail:
`l-gordon@northwestern.edu.
`
`© 2013 by American Society of Clinical
`Oncology
`
`0732-183X/13/3133w-4199w/$20.00
`
`DOI: 10.1200/JCO.2012.48.3685
`
`Disabling Immune Tolerance by Programmed Death-1
`Blockade With Pidilizumab After Autologous
`Hematopoietic Stem-Cell Transplantation for Diffuse Large
`B-Cell Lymphoma: Results of an International Phase II Trial
`
`Philippe Armand, Arnon Nagler, Edie A. Weller, Steven M. Devine, David E. Avigan, Yi-Bin Chen,
`Mark S. Kaminski, H. Kent Holland, Jane N. Winter, James R. Mason, Joseph W. Fay, David A. Rizzieri,
`Chitra M. Hosing, Edward D. Ball, Joseph P. Uberti, Hillard M. Lazarus, Markus Y. Mapara,
`Stephanie A. Gregory, John M. Timmerman, David Andorsky, Reuven Or, Edmund K. Waller,
`Rinat Rotem-Yehudar, and Leo I. Gordon
`See accompanying article on page 4268
`
`A
`
`B
`
`S
`
`T
`
`R
`
`A
`
`C
`
`T
`
`Purpose
`The Programmed Death-1 (PD-1) immune checkpoint pathway may be usurped by tumors,
`including diffuse large B-cell lymphoma (DLBCL), to evade immune surveillance. The reconstitut-
`ing immune landscape after autologous hematopoietic stem-cell transplantation (AHSCT) may be
`particularly favorable for breaking immune tolerance through PD-1 blockade.
`Patients and Methods
`We conducted an international phase II study of pidilizumab, an anti–PD-1 monoclonal antibody, in
`patients with DLBCL undergoing AHSCT, with correlative studies of lymphocyte subsets. Patients
`received three doses of pidilizumab beginning 1 to 3 months after AHSCT.
`Results
`Sixty-six eligible patients were treated. Toxicity was mild. At 16 months after the first treatment,
`progression-free survival (PFS) was 0.72 (90% CI, 0.60 to 0.82), meeting the primary end point.
`Among the 24 high-risk patients who remained positive on positron emission tomography after
`salvage chemotherapy, the 16-month PFS was 0.70 (90% CI, 0.51 to 0.82). Among the 35 patients
`with measurable disease after AHSCT, the overall response rate after pidilizumab treatment was
`51%. Treatment was associated with increases in circulating lymphocyte subsets including
`PD-L1E–bearing lymphocytes, suggesting an on-target in vivo effect of pidilizumab.
`Conclusion
`This is the first demonstration of clinical activity of PD-1 blockade in DLBCL. Given these results,
`PD-1 blockade after AHSCT using pidilizumab may represent a promising therapeutic strategy in
`this disease.
`
`J Clin Oncol 31:4199-4206. © 2013 by American Society of Clinical Oncology
`
`INTRODUCTION
`
`PD-1 (Programmed Death-1) is a member of the B7
`receptor family. Together with its ligands (PD-L1
`and PD-L2), it functions as an important check-
`point in the regulation of immune responses.1
`Those ligands are upregulated by the inflammatory
`environment and inhibit the function of PD-1–
`bearing lymphocytes. Thus the PD-1 immune
`checkpoint pathway serves to dampen peripheral
`lymphocyte activity in the context of inflammatory
`responses. This pathway seems to be co-opted by
`many tumors, preventing effective antitumor im-
`
`munity, and therefore represents a promising ther-
`apeutic target, as demonstrated in several solid
`tumor subtypes.2-5 Pidilizumab (CureTech, Yavne,
`Israel) is an anti–PD-1 humanized immunoglobulin
`G1 monoclonal antibody with preclinical antitumor
`activity in animal models.6-8 In a phase I trial in
`patients with advanced hematologic malignancies,
`pidilizumab showed a favorable safety profile and
`early evidence of clinical activity.9
`We conducted an international phase II study
`of pidilizumab in patients with diffuse large B-cell
`lymphoma (DLBCL) and primary mediastinal
`large B-cell lymphoma (PMBCL) after autologous
`
`© 2013 by American Society of Clinical Oncology
`Downloaded from jco.ascopubs.org on August 31, 2015. For personal use only. No other uses without permission.
`Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
`
`4199
`
`1 of 8
`
`BI Exhibit 1012
`
`
`
`Armand et al
`
`Table 1. Baseline Patient Characteristics (eligible patients)
`
`Table 1. Baseline Patient Characteristics (eligible patients) (continued)
`
`Variable
`
`Total No.ⴱ
`
`%
`
`Variable
`
`Total No.ⴱ
`
`Response to salvage therapy by PET
`Negative
`Positive
`PET not done
`Status before pidilizumab treatment#
`By CT imaging
`CR
`Not in CR
`By PET imaging
`Negative
`Positive
`PET not done/missing
`
`31
`24
`11
`
`31
`35
`
`45
`9
`12
`
`%
`
`47
`36
`17
`
`47
`53
`
`68
`18
`14
`
`57
`19-80
`
`25
`8-186
`
`No. of patients
`Age, years
`Median
`Range
`Race
`Asian
`Black
`White
`Hispanic
`Country
`Chile
`India
`Israel
`United States
`Sex
`Male
`Female
`Disease
`De novo DLBCL
`PMBCL
`Transformed indolent B-NHL
`IPI score at diagnosis
`0-1
`2
`3
`4-5
`Unknown
`Response to first-line therapy
`Complete remission
`Partial remission
`Stable or progressive disease
`Unknown
`Time from diagnosis to AHSCT, months
`Median
`Range
`Characteristics at relapse
`Stage†
`I
`II
`III
`IV
`Bulky‡
`Extranodal involvement§
`Marrow involvement储
`IPI¶
`0-1
`2
`3
`4-5
`No. of prior treatments
`1
`2
`3
`4
`Rituximab use
`With first-line therapy
`With salvage therapy
`With conditioning
`Radiation after transplantation
`(continued in next column)
`
`66
`
`7
`3
`52
`4
`
`1
`3
`7
`55
`
`43
`23
`
`49
`4
`13
`
`15
`11
`7
`7
`26
`
`45
`15
`5
`1
`
`6
`8
`9
`18
`18
`18
`29
`
`13
`7
`8
`2
`
`3
`47
`13
`3
`
`56
`54
`10
`5
`
`11
`5
`79
`6
`
`1
`4
`11
`83
`
`65
`35
`
`74
`6
`20
`
`23
`17
`11
`11
`39
`
`68
`23
`8
`1
`
`9
`12
`14
`27
`27
`27
`44
`
`20
`11
`12
`3
`
`5
`71
`20
`5
`
`85
`82
`15
`8
`
`Abbreviations: AHSCT, autologous hematopoietic stem-cell transplantation;
`B-NHL, B-cell non-Hodgkin lymphoma; CR, complete remission; CT, com-
`puted tomography; DLBCL, diffuse large B-cell lymphoma; IPI, International
`Prognostic Index; PET, positron emission tomography; PMBCL, primary
`mediastinal B-cell lymphoma; PR, partial remission; SD, stable disease.
`ⴱPercentages may not add to 100 because of rounding. Denominator used
`was all patients, including those with missing data.
`†Data missing on 25 patients.
`‡Data missing on nine patients.
`§Data missing on 10 patients.
`储Marrow biopsy was not performed on three patients; results were deter-
`mined at time of diagnosis for primary refractory patients (SD⫹PD) or at last
`relapse before transplant for the others.
`¶Data missing on 36 patients.
`#CT was required per protocol and used for eligibility determination; PET
`was obtained at the discretion of the treating clinician and not used for
`eligibility determination.
`
`hematopoietic stem-cell transplantation (AHSCT). PD-L1 is ex-
`pressed on suppressor immune cells in the tumor microenvironment
`and in at least a subset of DLBCL and PMBCL tumors,10-13 where it
`may alter the composition and function of tumor-infiltrating lym-
`phocytes,14 and therefore represents a valid therapeutic target.11,12
`Moreover, the post-AHSCT setting may be a particularly fertile con-
`text for PD-1 blockade. This is a state of low-volume residual disease,
`during which there is a remodeling of the immune system. Indeed, the
`majority of the circulating leukocytes in the first few months after
`AHSCT are natural killer cells, CD45RO⫹ memory/effector cells, and
`monocytes, which comprise pidilizumab’s target populations and
`whose presence in DLBCL tumors has been associated with a favorable
`prognosis.15-17 Therefore, PD-1 blockade early after AHSCT for pa-
`tients with DLBCL may prevent a tumor-dependent, PD-1 driven
`exhaustion of antitumor lymphocytes, leading to eradication of resid-
`ual disease and improvement in progression-free survival (PFS).
`
`PATIENTS AND METHODS
`
`Patients
`Patients 18 years and older could be consented for this study if they
`planned or had undergone AHSCT for DLBCL, PMBCL, or transformed
`indolent B-cell non-Hodgkin lymphoma. Only patients with chemotherapy-
`sensitive disease (at least partial remission18 after salvage therapy by computed
`tomography [CT] scans) were eligible. Confirmatory screening was performed
`between 30 and 90 days after AHSCT. To enroll onto the study and receive
`treatment, patients had to have CT scans before first drug administration
`showing no evidence of progressive disease (PD) from pretransplant assess-
`ment, as well as normal hematologic, renal, hepatic, and cardiac function.
`Patients with type 1 diabetes, immune deficiency, active autoimmune disease,
`
`4200
`
`JOURNAL OF CLINICAL ONCOLOGY
`© 2013 by American Society of Clinical Oncology
`Downloaded from jco.ascopubs.org on August 31, 2015. For personal use only. No other uses without permission.
`Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
`
`2 of 8
`
`BI Exhibit 1012
`
`
`
`PD-1 Blockade for DLBCL After AHSCT
`
`Table 2. Adverse Events
`
`Severity Grade
`
`All Grades
`
`1
`
`2
`
`3
`
`4
`
`5
`
`Event
`
`No. of
`Patients
`
`No. of
`Events
`
`No. of
`Patients
`
`No. of
`Events
`
`No. of
`Patients
`
`No. of
`Events
`
`No. of
`Patients
`
`No. of
`Events
`
`No. of
`Patients
`
`No. of
`Events
`
`No. of
`Patients
`
`No. of
`Events
`
`69
`19
`18
`14
`12
`12
`10
`9
`9
`
`613
`25
`21
`15
`19
`14
`15
`12
`12
`
`65
`3
`16
`9
`10
`11
`5
`8
`6
`
`392
`3
`19
`10
`14
`13
`6
`10
`8
`
`49
`7
`2
`5
`4
`1
`—
`2
`3
`
`145
`7
`2
`5
`5
`1
`
`2
`3
`
`Any AE
`Neutropenia
`Fatigue
`Upper respiratory tract infection
`Diarrhea
`Cough
`Thrombocytopenia
`Hyperglycemia
`Leukopenia
`Anemia
`Pyrexia
`Renal failure
`Vomiting
`Lymphopenia
`Cardiac arrest
`Duodenal ulcer
`GI hemorrhage
`General physical health decline
`Pain
`Clostridium difficile colitis
`Herpes zoster
`Lobar pneumonia
`Urinary tract infection
`Vascular injury
`Accident
`Fall
`Pelvic fracture
`Head injury
`Facial bone fracture
`aPTT prolonged
`Hypophosphatemia
`Bone pain
`Myositis
`Rhabdomyolysis
`Myelodysplastic syndrome
`Glioma
`Intracranial hemorrhage
`Subarachnoid hemorrhage
`Headache
`Tachypnea
`COPD
`ARDS
`Pneumothorax
`Hyperhidrosis
`Cholecystectomy
`DVT
`Hypertension
`Hypotension
`
`1
`
`1ⴱ
`—
`—
`—
`—
`—
`—
`—
`—
`
`1
`
`1
`
`9
`5
`—
`—
`—
`—
`2
`—
`—
`
`1
`
`1
`
`1†
`
`1
`1
`
`1
`
`14
`5
`
`4
`
`1
`
`1
`
`1
`
`1
`1
`
`1
`
`30
`9
`—
`—
`—
`—
`4
`—
`1
`3
`2
`2
`1
`1
`1
`1
`1
`1
`1
`1
`1
`1
`1
`1
`
`1
`1
`
`1
`1
`1
`1
`1
`1
`
`1
`
`1
`1
`1
`1
`
`1
`1
`1
`1
`1
`
`60
`10
`
`5
`
`1
`3
`2
`2
`2
`1
`1
`1
`1
`1
`1
`1
`1
`1
`1
`1
`
`1
`1
`
`1
`1
`1
`1
`1
`1
`
`1
`
`1
`1
`1
`1
`
`1
`1
`1
`1
`1
`
`NOTE. Data are shown as number of patients with a given AE and number of events. Only AEs representing ⱖ 2% of total events are shown for grade 1 and
`2 events.
`Abbreviations: AE, adverse event; aPTT, activated partial thromboplastin time; ARDS, acute respiratory distress syndrome; COPD, chronic obstructive pulmonary
`disorder; DVT; deep venous thrombosis.
`ⴱOne patient developed fatal disseminated zoster infection during the follow-up period.
`†One patient with pre-existing leukopenia and thrombocytopenia developed myelodysplasia 13 months after the last dose of pidilizumab. This was considered
`unrelated to study drug.
`
`www.jco.org
`
`© 2013 by American Society of Clinical Oncology
`Downloaded from jco.ascopubs.org on August 31, 2015. For personal use only. No other uses without permission.
`Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
`
`4201
`
`3 of 8
`
`BI Exhibit 1012
`
`
`
`Armand et al
`
`CNS involvement by lymphoma, active infection, other serious illness, con-
`current investigational treatment, or performance status more than 1 were
`excluded, as were pregnant or nursing patients.
`Patients were recruited at 30 centers in the United States, Israel, Chile,
`and India. All patients provided written informed consent. The study was
`approved by the offices for human research studies at the participating insti-
`tution and conducted in accordance with the principles of the Declaration of
`Helsinki. The study was supported by CureTech, and the data were analyzed by
`three of the authors (P.A., E.A.W., and L.I.G.) and by CureTech.
`Treatment and Monitoring
`Patients received treatment with pidilizumab administered intrave-
`nously at a dose of 1.5 mg/kg every 42 days for three cycles, beginning 30 to 90
`days from AHSCT. Premedication consisted of acetaminophen or ibuprofen,
`as well as diphenhydramine or promethazine. Patients were restaged with CT
`scans (with or without positron emission tomography [PET] scans, at the
`discretion of the treating clinicians) at confirmatory screening, then before the
`second and third cycles, and at 30, 44, and 69 weeks from the first day of
`treatment. Treatment was stopped if there was evidence of PD based on
`standard criteria.18 Patients were observed until 16 months from first pidili-
`zumab treatment, which corresponded to approximately 18 months from
`AHSCT. For patients with measurable disease at post-AHSCT screening, re-
`sponse to pidilizumab treatment was assessed18 according to the restaging
`schedule described previously, using the post-AHSCT measurements as the
`pretreatment baseline. Toxicity was graded using National Cancer Institute
`Common Terminology Criteria of Adverse Events v3.0.
`Correlative Studies
`Blood samples collected from all treatment sites from patients treated at
`least once with pidilizumab were analyzed by flow cytometry at two central
`laboratories of Esoterix LabCorp Services (Austin, TX) using study-specific
`validated methodologies. Forty-one prospectively specified leukocyte subsets
`based on cluster of differentiation marker expression were evaluated for abso-
`lute (per microliter) and relative numbers, as well as molecules of equivalent
`soluble fluorochrome (MESF). Validation studies for marker stability and
`inter- and intra-assay precision were conducted before initiating the tests in
`this study.
`Statistical Considerations
`The primary end point of this study was the 16-month progression-free
`proportion from the time of first pidilizumab administration among all eligi-
`ble patients who received at least one dose of pidilizumab. Secondary end
`points included safety and toxicity, PFS, and overall survival (OS); immuno-
`genicity of pidilizumab; and immune subset analyses. OS was defined as the
`time from first treatment to death, and PFS as the time to death, relapse, or
`
`progression. OS and PFS were calculated using the Kaplan-Meier method. On
`the basis of data available at the time of study design, the 18-month PFS after
`transplantation for chemosensitive patients was estimated to be approximately
`60% to 65%.19,20 This time corresponds to approximately 16 months from the
`planned start of pidilizumab on this trial. Because the number of patients
`enrolled was anticipated to range from 64 to 80 patients, an observed 16-
`month PFS from start of pidilizumab of at least 69% was considered to warrant
`further study. This design had at least 87% probability of concluding the
`treatment promising if the true 16-month PFS was 75% and less than 10%
`probability if the true 16-month PFS was 60% (given the exact binomial
`distribution applied to all possible sample sizes between 64 and 80 patients).
`Patients were eligible if they met all eligibility criteria and received at least one
`dose of pidilizumab.
`For exploratory measurements of changes in immune subsets, we com-
`pared absolute numbers of prespecified circulating lymphocytes of a given
`immunophenotype before the first treatment and at 24 hours, 6, 12, and 16
`weeks afterwards. We also measured MESF to assess for change in surface
`expression of selected markers. Pre- and post-treatment values were compared
`using paired Wilcoxon signed rank testing for individual time points, adjusted
`for multiple comparisons, as well as repeated measures analysis using a log10
`transformation (with SAS proc mixed). All P values are two-tailed, using a
`threshold for statistical significance of .05 except as noted. The data were
`analyzed using SAS version 9.2 (SAS Institute, Cary, NC).
`
`RESULTS
`
`Patients
`Patients could sign consent before or after AHSCT, but were
`required to pass confirmatory screening to be enrolled onto the study
`and treated. Among the 97 patients who gave consent, 25 were screen
`failures at confirmatory screening (including six with PD, five on
`concurrent disallowed treatment, four who withdrew consent, two
`with CNS disease, and two with infection). Therefore, 72 patients
`received at least one dose of pidilizumab (treated subset) at a median
`of 2.6 months after AHSCT (range, 1.1 to 4.1 months). Sixty patients
`(83%) completed all three cycles. Ten patients withdrew from the
`study before the 16-month follow-up visit for reasons other than
`death or progression: loss to follow-up (n ⫽ 2), investigator decision
`(n ⫽ 3), withdrawal of consent (n ⫽ 3), protocol violation (n ⫽ 1),
`and adverse event (AE; n ⫽ 1). On final review, six patients were
`
`OS
`PFS
`
`4
`8
`12
`Time Since First Treatment (months)
`
`16
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`Survival (%)
`
`B
`
`OS
`PFS
`
`4
`8
`12
`Time Since First Treatment (months)
`
`16
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`A
`
`Survival (%)
`
`Fig 1. Progression-free survival (PFS) and overall survival (OS) after pidilizumab treatment. (A) PFS and OS of all eligible patients. (B) PFS and OS of the 24 eligible
`patients who remained positive on positron emission tomography after salvage therapy.
`
`4202
`
`JOURNAL OF CLINICAL ONCOLOGY
`© 2013 by American Society of Clinical Oncology
`Downloaded from jco.ascopubs.org on August 31, 2015. For personal use only. No other uses without permission.
`Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
`
`4 of 8
`
`BI Exhibit 1012
`
`
`
`PD-1 Blockade for DLBCL After AHSCT
`
`Total CD14+
`CD14+ PD-L1+
`CD14+ PD-L2+
`Total CD4+ CD25+
`CD4+ CD25+ PD-L1+
`
`24 hours
`
`6 weeks
`
`12 weeks
`Time
`
`16 weeks
`
`CD4+ CD62L+ CD127+
`CD4+ CD62L- CD127+
`CD8+ CD62L+ CD127+
`CD8+ CD62L- CD127+
`
`24 hours
`
`6 weeks
`
`12 weeks
`Time
`
`16 weeks
`
`PD-L1 on CD14+
`PD-L2 on CD14+
`PD-L1 on CD4+ CD25+
`CD127 on CD4+ CD62L+
`CD127 on CD4+ CD62L-
`CD127 on CD8+ CD62L+
`CD127 on CD8+ CD62L-
`
`24 hours
`
`6 weeks
`
`12 weeks
`Time
`
`16 weeks
`
`200
`180
`160
`140
`120
`100
`80
`60
`40
`20
`0
`
`350
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`-20
`
`A
`
`Mean Change From Baseline
`
`No. of Cells/μL (%)
`
`B
`
`Mean Change From Baseline
`
`No. of Cells/μL (%)
`
`C
`
`Mean Change From Baseline
`
`No. of Cells/μL (%)
`
`Fig 2. Changes in absolute number of circulating lymphocyte subsets and
`surface marker expression after pidilizumab administration. (A) Changes in
`circulating number of PD-L1 (B7-H1) and PD-L2 (B7-DC) –positive monocytes
`and T cells.
`(B) Changes in circulating number of peripheral and central
`memory CD8 T cells. (C) Changes in expression of selected surface markers
`on monocytes and T cells.
`
`pidilizumab resulted in a significant increase in the absolute number
`of PD-L1–bearing activated helper T cells (CD4⫹ CD25⫹ PD-L1⫹),
`apparent 24 hours after first treatment and sustained until at least 16
`weeks (Fig 2A). There also seemed to be some changes in PD-1 ligand-
`bearing monocytes (CD14⫹ PD-L1⫹ and CD14⫹ PD-L2⫹ cells). For
`the latter subsets, the mean percentage increase was larger than the
`
`determined ineligible (four because of refractoriness to salvage ther-
`apy, one for PD on the first day of treatment, and the other for
`disallowed concomitant treatments). The baseline characteristics of
`the 66 eligible patients are shown in Table 1. At the time of post-
`AHSCT restaging, 47% of patients were in complete remission (CR)
`by CT. Fifty-five patients had a PET scan after salvage; 31 (47%) were
`in PET-CR. Fifty-four patients had a post-AHSCT PET scan; 45 (68%)
`were in PET-CR at that time.
`
`Safety and Toxicity
`Among all 72 treated patients, a total of 613 AEs occurred in 69
`(96%) of patients (Table 2), among which 135 were considered related
`to treatment. The most frequently reported grade 3 to 4 AEs were
`neutropenia (19% of patients) and thrombocytopenia (8%). All pa-
`tients with grade 4 neutropenia responded to growth factor treatment
`and remained asymptomatic. One patient died of disseminated herpes
`zoster 10 months after the third dose of pidilizumab, which was
`considered unrelated to study treatment. Twenty-three patients
`(32%) experienced at least one serious AE, and three patients (4%)
`experienced a related serious AE. There was no evidence of significant
`autoimmune toxicity, no infusion reactions, and no treatment-
`related mortality.
`
`Clinical Outcome
`Among the 66 eligible patients, 18 experienced disease progres-
`sion or died before the 16-month time point. The 16-month PFS from
`first treatment (the primary end point) was 0.72 (90% CI, 0.60 to 0.82;
`Fig 1A). The study therefore met its primary end point. Nine patients
`died during the study period between 2.3 and 15.3 months; the cause
`of death was lymphoma in eight patients and disseminated herpes
`zoster in one patient. The 16-month OS for eligible patients was 0.85
`(90% CI, 0.74 to 0.92; Fig 1A). Among the 24 patients who remained
`PET-positive at the conclusion of salvage therapy, 16-month PFS was
`0.70 (90% CI, 0.51 to 0.82; Fig 1B). Among the 31 PET-negative
`patients, 16-month PFS was 0.72 (90% CI, 0.56 to 0.84); among the 11
`patients who did not have a postsalvage PET scan, 16-month PFS was
`0.72 (90% CI, 0.42 to 0.88). No significant difference was detected in
`the PFS or OS between patients when stratified by disease status
`assessed by CT scans after AHSCT, age, time to first relapse, time from
`diagnosis to AHSCT, or salvage regimen; however, there was limited
`power for those comparisons. We also performed an intent-to-treat
`analysis for the 72 treated patients. The PFS at 16 months from first
`pidilizumab treatment in this cohort was 0.68 (90% CI, 0.59 to 0.77),
`and OS was 0.84 (90% CI, 0.77 to 0.91).
`Among the 35 eligible patients with measurable disease at
`screening post-AHSCT and before the first dose of pidilizumab, 12
`(34%) achieved a CR by CT criteria after pidilizumab treatment,
`and six (17%) achieved a partial remission (PR), for an overall
`response rate of 51%. In addition, 13 patients (37%) had stable
`disease, whereas four (11%) had PD. The median time to docu-
`mented response was 30 weeks (range, 6 to 69 weeks). Among the
`nine patients who had residual disease after AHSCT and a positive
`PET scan, the overall response rate was 33%, and an additional
`44% had stable disease.
`
`Immune Subset Analyses
`Figure 2 and Table 3 detail changes in selected lymphocyte sub-
`sets and marker expression among eligible patients. Treatment with
`
`www.jco.org
`
`© 2013 by American Society of Clinical Oncology
`Downloaded from jco.ascopubs.org on August 31, 2015. For personal use only. No other uses without permission.
`Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
`
`4203
`
`5 of 8
`
`BI Exhibit 1012
`
`
`
`Armand et al
`
`Table 3. Changes in Selected Lymphocyte Subsets and Marker Expression
`
`24 Hours
`
`6 Weeks
`
`12 Weeks
`
`16 Weeks
`
`Variable
`
`Adjusted
`P
`
`Mean
`Change
`
`Median
`Change
`
`P
`
`Mean
`Change
`
`Median
`Change
`
`P
`
`Mean
`Change
`
`Median
`Change
`
`P
`
`Mean
`Change
`
`Median
`Change
`
`P
`
`.2
`.3
`.3
`< .0001
`.0004
`.0025
`.07
`.003
`.0002
`
`Lymphocyte population
`CD14⫹
`CD14⫹ PD-L1⫹
`CD14⫹ PD-L2⫹
`CD4⫹CD25⫹
`CD4⫹CD25⫹ PD-L1⫹
`CD4⫹CD62L⫹ CD127⫹
`CD4⫹CD62L– CD127⫹
`CD8⫹CD62L⫹ CD127⫹
`CD8⫹CD62L– CD127⫹
`Marker expression
`PD-L1 on CD14⫹
`.6
`PD-L2 on CD14⫹
`.2
`PD-L1 on CD4⫹ CD25⫹
`.04
`CD127 on CD4⫹CD62L⫹
`.0001
`CD127 on CD4⫹CD62L–
`.0001
`CD127 on CD8⫹CD62L⫹ < .0001
`CD127 on CD8⫹CD62L–
`.7
`
`⫹0.00
`⫹6.9
`⫺9.65
`⫹47.6
`⫺8.82
`⫹22.6
`⫹7.77
`⫹7.7
`⫹80.4 ⫹29.17
`⫹72.6
`⫹9.97
`⫹12.8
`⫹6.83
`⫹34.6 ⫹25.93 < .0001
`⫹27.9 ⫹11.04
`.03
`
`.015
`.041
`.003
`.02
`
`⫹0.0
`⫹14.4
`⫹53.3 ⫹10.0
`⫹61.4 ⫹17.3
`⫹5.9
`⫹0.0
`⫹87.1
`⫹0.0
`⫹134.6
`-5.5
`⫹18.1
`⫹6.5
`⫹29.4
`⫹9.7
`⫹55.3
`⫹2.3
`
`⫺5.3
`⫹9.3
`⫹1.0 ⫺20.2
`⫺6.7
`⫹60.5
`⫹70.7 ⫺31.3
`⫹75.6 ⫹12.1
`⫹86.7
`⫺1.6
`.09
`⫹50.7 ⫹30.0
`⫹16.7
`⫺2.6
`.0001
`⫹119.9 ⫹50.0 < .0001
`.013 ⫹137.4 ⫹16.7
`⫹346.7 ⫹24.2
`⫹192.8
`⫺2.9
`.0005
`⫹32.8 ⫹19.8
`⫹23.6
`⫹6.3
`.002
`⫹119.8 ⫹37.1
`⫹52.2
`⫹9.0
`.0002
`⫹88.8 ⫹54.9 < .0001 ⫹108.1 ⫹22.5
`.001
`
`.02
`.08
`.01
`
`.006
`
`⫹0.0
`⫹8.8
`⫺1.0
`⫹0.6
`⫹5.0
`⫹6.9
`⫹15.9 ⫹15.9
`⫹18.9
`⫹9.3
`⫹8.8
`⫹6.2
`⫹1.8
`⫺2.9
`
`⫹6.5
`⫹0.3
`⫹6.5
`⫹14.0
`⫹20.8
`⫹29.4
`⫹0.7
`
`⫹2.75
`⫺1.14
`⫹4.58
`⫹3.69
`⫹6.39
`⫹7.33
`⫹0.04
`
`.05
`.05
`.036
`
`⫹1.0
`⫹11.4
`⫺5.1
`⫺0.5
`⫹8.4
`⫹7.3
`⫹27.9 ⫹21.4
`⫹40.7 ⫹20.6
`⫹68.0 ⫹24.4
`⫹1.0
`⫺0.9
`
`.09
`
`⫹4.6
`⫹14.1
`⫺6.8
`⫺2.2
`⫹18.2 ⫹15.1
`⫹39.1 ⫹28.5 < .0001
`⫹58.2 ⫹33.0 < .0001
`⫹81.6 ⫹37.0 < .0001
`⫹3.6
`⫹1.5
`
`.003
`.001
`.0007
`
`.0007
`.004
`
`NOTE. Changes are reported compared with baseline values as both the mean and median percentage change in the absolute number of the selected subset or
`marker. Unadjusted P values are based on Wilcoxon signed rank testing; only P values ⬍ .1 are reported. Bolded P values indicate significance after adjustment for
`multiple comparisons (P ⬍ .0125). Adjusted P values test the significance of changes over time using repeated measurement analysis (see Patients and Methods).
`
`median increase at all time points (Table 3), suggesting that pidili-
`zumab induced large increases in those cells that were early and
`sustained, but restricted to a subset of patients. MESF analysis
`excluded the possibility that these changes could be solely ex-
`plained by upregulation of surface markers. There were also signif-
`icant increases in the absolute number of circulating CD8⫹
`peripheral (CD62L-CD127⫹) and central (CD62L⫹CD127⫹)
`memory T cells, as well as in CD4⫹ central memory T cells (Fig 2B).
`Finally, we found an increase in the cell surface expression of the
`interleukin 7␣receptor (CD127) on peripheral and central mem-
`ory CD4⫹ and CD8⫹ T cells (Fig 2C).
`
`DISCUSSION
`
`Monoclonal antibody therapy as a means of targeting immune
`checkpoints has emerged as a viable and effective antitumor
`strategy.3-5,9,21,22 Hematologic malignancies may be particularly at-
`tractive targets for this type of treatment, as patients with even ad-
`vanced myeloid or lymphoid tumors can be cured by adoptive
`immunotherapy delivered in the context of allogeneic HSCT. This
`raises the possibility that patients’ own immune systems can be har-
`nessed to eradicate those diseases, if the mechanisms that lead to
`immune tolerance of the tumor can be safely disabled. In the present
`trial, we show that the anti–PD-1 monoclonal antibody pidilizumab
`can be safely given to patients with DLBCL after AHSCT. The lack of
`significant autoimmune toxicity in our trial stands in contrast to the
`clinical experience so far with cytotoxic T-lymphocyte antigen-4
`blockade.23 Treatment was associated with an apparent CR rate of
`34% and overall response rate of 51% among patients with measur-
`able disease after transplant. This suggests direct antitumor activity,
`although given this study’s design, we cannot rule out the possibility
`
`that the residual radiologic abnormalities in some cases reflected
`treated disease or inflammation rather than the presence of viable
`lymphoma. Moreover, with a 16-month PFS of 0.72, the study met its
`prespecified primary end point. These results may compare favorably
`with those of two recent multicenter randomized clinical trials in this
`population, although our exclusion of patients who experienced re-
`lapse early after AHSCT precludes a direct comparison.24,25 In the
`rituximab era, as the prognosis of patients with relapsed or refractory
`disease is worse,24 new therapies are needed to increase the efficacy of
`salvage and to increase the success rate of AHSCT in patients without
`a PET-CR at transplantation26-28 who have a poorer outcome. We
`recently reported the outcomes of 105 patients with DLBCL who
`underwent transplantation in the last decade at our own institutions.29
`Among this cohort, we examined (on an institutional review board–
`approved study) the outcome of the 46 patients who were chemosen-
`sitive but PET positive after salvage and who would have otherwise
`met the eligibility criteria for the present study, including no progres-
`sion or relapse within 2 months of AHSCT. In this group, the 18-
`month post-AHSCT PFS was 0.52 (90% CI, 0.39 to 0.63). In the
`present study, the PFS was 0.72 among PET-positive patients treated
`with pidilizumab. This compares favorably with our historical experi-
`ence and with other published cohorts.26-28,30-32 Although not all
`patients on the present study had a postsalvage PET (which was not
`mandated per protocol), the absence of an apparent difference in
`outcome between those who did and those who did not argues against
`a strong selection bias, again with the caveat of the small numbers
`involved. We emphasize that the study was not powered for a com-
`parison among PET subgroups or with historical controls. Nonethe-
`less, our findings support the hypothesis that PD-1 blockade may be a
`viable therapeutic strategy in the high-risk subset of patients with
`residual disease and may overcome the negative prognostic value of a
`
`4204
`
`JOURNAL OF CLINICAL ONCOLOGY
`© 2013 by American Society of Clinical Oncology
`Downloaded from jco.ascopubs.org on August 31, 2015. For personal use only. No other uses without permission.
`Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
`
`6 of 8
`
`BI Exhibit 1012
`
`
`
`PD-1 Blockade for DLBCL After AHSCT
`
`pretransplant positive PET scan. This hypothesis should now be tested
`in a randomized clinical trial.
`The increase in the number of some PD-1 ligand-bearing lym-
`phocyte subsets within 24 hours of drug infusion is consistent with an
`on-target effect of pidilizumab. Indeed, the interaction of PD-1 with
`either of its cognate ligands seems to induce cell death and signal
`suppression in lymphocytes and monocytes.33 The increase in circu-
`lating levels of cells expressing PD-1 ligands may therefore reflect the
`reversal of PD-1’s inhibition of cell survival or proliferation, although
`the rapidity of some of the changes occurring after treatment suggests
`that blockade of apoptosis is not the main or only mechanism at play
`and that mobilization of those cells from their reservoirs may be
`important. Also, our results suggest significant variability in the
`changes among different patients, implying additional complexity in
`the immune effects of PD-1 blockade in the setting of a reconstituting
`immune system. We also found an increase in effector and peripheral
`memory cell subsets, consistent with in vitro data that pidilizumab
`enhances the survival of human CD4⫹CD45RO⫹ cells (CureTech,
`unpublished data) within 72 hours of treatment. The increase in the
`cell surface expression of the interleukin 7␣receptor CD127, pivotal
`for the maturation and survival of memory T cells, suggests that
`pidilizumab induces molecular events associated with the fate of
`specific memory T-cell subsets. These analyses are only explor-
`atory, reported for hypothesis-generating purposes, and should be
`prospectively validated.
`Because the expression of PD-L1 on DLBCL cells may be re-
`stricted to a subset of tumors,10 it may be that future selection of
`patients for PD-1 blockade on the basis of ligand expression in the
`tumor or microenvironment could lead to a greater clinical benefit in
`the appropriate patient subgroups; this could not be ascertained on
`this study because we did not have access to tumor material for most
`patients. Furthermore, the ability to evaluate for on-target effects that
`predict outcome could provide a method to adapt post-AHSCT ther-
`apy in those patients. Although the answers to those questions must
`await future prospective trials, the present study represents the first
`efficacy trial of immune checkpoint blockade in hematologic malig-
`nancies. Our rapidly expanding scientific knowledge in this area, cou-
`pled with the availability of a growing number of monoclonal
`antibodies targeting those pathways, will doubtlessly lead into broader
`investigations of this strategy in other settings.
`
`AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS
`OF INTEREST
`
`Although all authors completed the disclosure declaration, the following
`author(s) and/or an author’s immediate family member(s) indicated a
`financial or other interest that is relevant to the subject matter under
`consideration in this article. Certain relationships marked with a “U” are
`those for which no compensation was received; those relationships marked
`with a “C” were compensated. For a detailed description of the disclosure
`categories, or for more information about ASCO’s conflict of interest policy,
`please refer to the Author Disclosure Declaration and the Disclosures of
`Potential Conflicts of Interest section in Information for Contributors.
`Employment or Leadership Position: Rinat Rotem-Yehudar, Cure Tech
`(C) Consultant or Advisory Role: Philippe Armand, Merck (C); Arnon
`Nagler, CureTech (C); Edie A. Weller, CureTech (C); Yi-Bin Chen,
`Otsuka (C), Genzyme (C); David A. Rizzieri, Teva (C), Celgene (C),
`Celgene (C), Novartis (C), Spectrum (C); Leo I.