of 16 /16
Anti-CD27 Agonist Antibody Varlilumab with Nivolumab for Colorectal and Ovarian Cancer: Phase 1/2 Clinical Trial Results Rachel E. Sanborn 1 , Michael J. Pishvaian 2 , Margaret K. Callahan 3 , Amy Weise 4 , Branimir I. Sikic 5 , Osama Rahma 6 , Daniel Cho 7 , Naiyer Rizvi 8 , Jose Lutzky 9 , Rhonda L. Bitting 10 , Alexander Starodub 11 , Antonio Jimeno 12 , Michael Yellin 13 , Tracey Rawls 13 , Laura Vitale 13 , Abdel Halim 13 , Hui Zhang 13 and Tibor Keler 13 1 Rachel E. Sanborn 1 Robert W. Franz Cancer Research Center, Earle A. Chiles Research Institute, Providence Cancer Center, Portland, OR, 2 Georgetown University, Washington, DC, 3 Memorial Sloan Kettering Cancer Center, New York, NY, 4 Karmanos Cancer Institute, Detroit, MI, 5 Stanford Medical Center, Stanford, CA, 6 Dana Farber Cancer Institute, Boston, MA, 7 Laura & Isaac Perlmutter Cancer Center, NY, NY, 8 Columbia University Medical Center, NY, NY, 9 Mount Sinai Comprehensive Cancer Center, Miami Beach, FL, 10 Wake Forest Baptist Heath, Winston-Salem, NC, 11 Parkview Research Center, Fort Wayne, IN, 12 University of Colorado Cancer Center, Aurora CO, 13 Celldex Therapeutics, Hampton, NJ

Anti-CD27 Agonist Antibody Varlilumab with Nivolumab for ...Anti-CD27 Agonist Antibody Varlilumab with Nivolumab for Colorectal and Ovarian Cancer: Phase 1/2 Clinical Trial Results

  • Author
    others

  • View
    0

  • Download
    0

Embed Size (px)

Text of Anti-CD27 Agonist Antibody Varlilumab with Nivolumab for ...Anti-CD27 Agonist Antibody Varlilumab...

  • Anti-CD27 Agonist Antibody Varlilumab with Nivolumabfor Colorectal and Ovarian Cancer:

    Phase 1/2 Clinical Trial Results

    Rachel E. Sanborn1, Michael J. Pishvaian2, Margaret K. Callahan3, Amy Weise4, Branimir I. Sikic5, Osama Rahma6, Daniel Cho7, Naiyer Rizvi8, Jose Lutzky9, Rhonda L. Bitting10, Alexander Starodub11, Antonio Jimeno12, Michael Yellin13, Tracey Rawls13, Laura Vitale13, Abdel Halim13,

    Hui Zhang13 and Tibor Keler13

    1Rachel E. Sanborn

    1 Robert W. Franz Cancer Research Center, Earle A. Chiles Research Institute, Providence Cancer Center, Portland, OR, 2 Georgetown University, Washington, DC, 3 Memorial Sloan Kettering Cancer Center, New York, NY, 4 Karmanos Cancer Institute, Detroit, MI, 5Stanford Medical Center, Stanford, CA, 6Dana Farber Cancer Institute, Boston, MA, 7Laura & Isaac Perlmutter Cancer Center, NY, NY, 8Columbia University Medical Center, NY, NY, 9Mount Sinai Comprehensive Cancer Center, Miami Beach, FL, 10Wake Forest Baptist Heath, Winston-Salem, NC, 11Parkview Research Center, Fort Wayne,

    IN, 12University of Colorado Cancer Center, Aurora CO, 13Celldex Therapeutics, Hampton, NJ

  • Combo CD27 PD-(L)1

    DNA replication genes

    Cytotoxicity andProliferation genes

    Combination of CD27 Costimulation with PD-(L)1 Blockade

    1 Buchan. et al. Clin Can Res 20182 Bullock, et al. SITC 20143 Infante, et al. ASCO 20144 Burris, et al. JCO 2017

    BCL1 lymphoma

    Cooperative roles of anti-CD27 and PD-(L)1 drives proliferation and cytotoxicity of T cells1

    Adoptive transfer of Pmel-TgT cells stimulated in vivo with

    peptide and Abs

    Combining CD27 agonist Abs with PD-(L)1 blockade improves antitumor responses in several preclinical models1

    CD27: Member of the TNF-receptor superfamily • Single ligand is CD70 (tightly regulated)• Constitutively expressed on most T cells/subset of B and NK cells1

    • CD27 activation:– Signaling through Traf2, Traf 5– Activation of the NF-κB pathway– Cell survival, activation, proliferation– Role in generation and long-term maintenance of T cell immunity– Role in NK cell differentiation/activation

    Varlilumab: Fully human IgG1 CD27 agonist mAb• Well tolerated as single agent, no MTD identified• Single-agent antitumor activity demonstrated in advanced,

    refractory solid tumors or hematologic malignancies (n=90) 2, 3, 4

    (varlilumab)

    2Rachel E. Sanborn

  • Phase 2 Cohorts*Nivolumab flat dose (240 mg) q 2 weeks

    CRC (n=18)**Varlilumab3 mg/kg q 2 weeks

    ** Planned enrollment

    Phase 1/2 Study of Varlilumab in Combination with Nivolumab

    Colorectal and Ovarian Cancer ExperienceActual Enrollment Status

    (cut-off 13Apr18)Phase 1 Phase 2 Overall

    Ovarian Cancer 8 58 66 7 continue treatment

    CRC 21 21 42 2 continue treatment

    Primary objective: Estimation of objective response rate (ORR)ORR observed with nivolumab monotherapy:

    CRC: MSI− -low 0-5%2, MSI-high 31%3Ovarian: − 6-15%4,5

    Additional objectives: PFS, OS, Immunogenicity, PK, PD

    CNS, central nervous system; CRC, colorectal cancer; MTD, maximum tolerated dose; MSI, microsatellite instability; OS, overall survival; PD, pharmacodynamics; PFS, progression free survival; PK, pharmacokinetics

    3Rachel E. Sanborn

    Varlilumab3 mg/kg q 2 weeks (n=18)3 mg/kg q 12 weeks (n=18)0.3 mg/kg q 4 weeks (n=18)

    Ovarian (n=54)**

    0

    20

    40

    60

    80

    100

    0 14 28 42 56 70 84 98 1120

    20

    40

    60

    80

    100

    0 14 28 42 56 70 84 98 1120

    20

    40

    60

    80

    100

    0 14 28 42 56 70 84 98 112

    0.3 mg/kg Q4wk3 mg/kg Q2wk 3 mg/kg Q12wk

    Days

    Varl

    ilum

    ab, µ

    g/m

    l

    Predicted PK: Alternate dosing regimens/Intermittent CD27 signaling

    Key Eligibility Criteria• Progressive, recurrent or refractory ovarian

    cancer, CRC, SCCHN, melanoma, or NSCLC• No prior anti-PD-(L)1 • ≥ 3 month washout for T cell directed mAbs

    (inc. anti-CTLA-4) • ≤ 5 prior regimens for advanced disease• No active CNS metastases • No autoimmune disease• CRC: Progression or intolerant to

    fluoropyrimidine, oxaliplatin, and irinotecan plus bevacizumab, cetuximab or panitumumab (if KRAS wild type), and regorafenib.

    • Ovarian cancer: platinum-taxane frontline therapy

    * Varlilumab administered for up to ~32 weeks (q2w and q4w schedules) or ~48 weeks (q12w schedule); nivolumab continues until progression

    Phase 1 Dose Escalation/Expansion*

    Nivolumab 3 mg/kg q 2 weeksVarlilumab escalating doses

    q 2 weeks:

    • Well tolerated, MTD not identified1• No clear varlilumab dose

    response1

    0.1 mg/kg (n=6)

    10 mg/kg (n=15)

    1 mg/kg (n=15)

    1. Sanborn, ASCO 20172. Overman, ASCO 20163. Overman, Lancet 2017

    4. Brahmer, NEJM 20125. Hamanishi, JCO 2015

  • Baseline Patient Characteristics

    ECOG, Eastern Cooperative Oncology Group; MSI, microsatellite instability; N/A, not applicable1 Denominator represents patients with tumor assessed for PD-L1 status. PD-L1+ criteria: ≥ 1% tumor cells staining positive, using the BMS

    developed PD-L1 IHC method at a central lab

    Ovarian Cancer (n=66)

    CRC (n=42)

    Age, years (median [range]) 64 (40-89) 55 (29-76)Female (n [%]) 66 (100%) 16 (38%)ECOG performance status (n [%])

    0 22 (33%) 18 (43%)1 44 (67%) 24 (57%)

    Stage IV at study entry (n [%]) 60 (91%) 42 (100%)No. of prior treatment regimens (median [range]) 3 (1-8) 4 (1-9)

    Immunotherapy/cytokine 2 (3%) 5 (12%)

    PD-L1+ tumor (n/n [%])1 20/59 (34%) 5/38 (13%)MSI Status (n [%])

    MSI-High N/A 1 (2%)MSI-Low or MMR proficient N/A 21 (50%) MSI unknown N/A 20 (48%)

    4Rachel E. Sanborn Data cut-off as of April 13, 2018

  • • No evidence of additive toxicity for the combination

    • Toxicity profile similar across varlilumab dosing regimens

    • 10% of patients with CRC and ovarian cancer discontinued study treatment due to toxicity

    • 10 patients with treatment-related SAEs

    • One treatment-related death (pneumonitis in a patient with pulmonary metastases and prior history of chemotherapy-induced pneumonitis)

    Varlilumab & Nivolumab Combination Therapy is Well Tolerated

    Data shown as N (%). Table includes adverse events assessed as related to either varlilumab or nivolumab for > 10% of patients overall and treatment-related SAEs (*)

    Treatment-Related Adverse Events Ovarian Cancer (n=66) CRC (n=42)All Grade

    3-4Grade

    5All Grade

    3-4Grade

    5Infusion Reaction 11 (17) 0 0 13 (31) 0 0Pruritus 12 (18) 0 0 6 (14) 0 0Rash 12 (18) 0 0 6 (14) 0 0Fatigue 5 (8) 0 0 5 (12) 0 0Rash maculo-papular 8 (12) 0 0 4 (10) 1 (2) 0Lymphopenia 4 (6) 2 (3) 0 7 (17) 5 (12) 0Nausea 4 (6) 0 0 7(17) 0 0ALT increased* 4 (6) 1 (2) 0 2 (5) 1 (2) 0Lipase increased * 4 (6) 4 (6) 0 1 (2) 1 (2) 0Abdominal pain * 3 (5) 1 (2) 0 0 0 0Acute kidney injury * 3 (5) 2 (3) 0 0 0 0Pneumonitis * 0 0 0 2 (5) 0 1 (2)Tumor lysis syndrome * 1 (2) 1 (2) 0 0 0 0Hepatitis * 1 (2) 1 (2) 0 0 0 0Colitis * 1 (2) 1 (2) 0 0 0 0Small intestinal obstruction * 1 (2) 1 (2) 0 0 0 0Diarrhea * 4 (6) 0 0 5 (12) 0 0Peripheral sensory neuropathy * 1 (2) 0 0 0 0 0

    5Rachel E. Sanborn

  • Tumor Response: Ovarian Cancer

    Nivolumab dosing: 3 mg/kg q2w in Phase 1 and 240 mg q2w in Phase 2 ORR, Objective response rate; DCR, Disease control rate (best response of SD or better for ≥ 3 months). Analyses based on response-evaluable population (includes patients with symptomatic deterioration or death in absence of post-treatment tumor assessment).

    Rachel E. Sanborn 6

    Best Response: Partial Response (PR)Single time-point PR (uPR)Stable Disease (SD)Progressive Disease (PD) Not Evaluable (NE)

    PD-L1 Status: Negative

    Numerical values represent % cells positive

    Patient continues treatment-* Includes data provided after the analysis cut-off date

    Response rate by PD-L1 status: PDL-1 positive: 20% (n=4 of 20; 3 PR, 1 uPR)PDL-1 negative: 14% (n=5 of 37; 4 PR, 1 uPR)

  • Tumor Response: CRC

    7

    Nivolumab dosing: 3 mg/kg q2w in Phase 1 and 240 mg q2w in Phase 2 ORR, Objective response rate; DCR, Disease control rate (best response of SD or better for ≥ 3 months). Analyses based on response-evaluable population (includes patients with symptomatic deterioration or death in absence of post-treatment tumor assessment)

    Nivolumab + Varlilumab q2w

    0.1-10 mg/kg 3 mg/kg

    ORR: 1/21 (5%) 1/20 (5%)

    DCR: 4/21 (19%) 4/20 (20%)

    -100%

    -75%

    -50%

    -25%

    0%

    25%

    50%

    75%

    100%

    Best Response: Partial Response (PR)Single time-point PR (uPR)Stable Disease (SD)Progressive Disease (PD) Not Evaluable (NE)

    PD-L1 Status: Negative

    Numerical values represent % cells positive

    Patient continues treatment

    -

    Max

    imum

    % S

    hrin

    kage

    -

    - -

    - - 3 - - - - - -- -

    -- -- - -- - -2 - - - --

    1 mg/kgMMR-proficient

    MSI high

    Phase 1 Phase 2

    0

    500

    1000

    1500

    2000

    2500

    1004 2018 2019

    Synonymous

    Non-synonymous

    CRC pt: PR PD PD

    Mut

    atio

    ns

    Mutational Load

    70-

    Rachel E. Sanborn

    Molecular analysis on baseline tumor • Tissue from 2 patients with progressive disease used for comparison• Strong pattern of differentially expressed genes

    • Similar expression of DNA repair enzymes• High mutational burden likely contributed to response

    • May be result of mutations in MLH-1 and MSH-6

    CRC pt: PR PD PD

    Differentially ExpressedGenes

    Patient with CRC initially considered MMR-proficientExperienced near• -CR (95% tumor shrinkage), continues at 35 monthsPath IHC report: PMS• 2, hMLH-1, hMSH-2, hMSH-6 all present

  • Analysis of Immune Monitoring and Correlations with Outcome Immune Monitoring Parameters:• Peripheral blood: serum factors by multiplex; flow cytometry on whole blood/PBMC• Tumor biopsies: baseline and on-treatment (day 29) immunohistochemistry• Molecular profiling (in progress)

    Benefit No BenefitCorrelative analysis:• Patients with ovarian cancer sorted by:

    − “Benefit” = SD ≥ 16wks, uPR, PR or CR − “No Benefit” = PD and SD < 16 wks− Correlative analysis included all ovarian

    patients • Patients with CRC; Too few “Benefit” for

    analysis

    8Rachel E. Sanborn

    0 1 6 3 2 4 8 6 4

    - 1 0 0

    - 5 0

    0

    5 0

    1 0 0

    W e e k

    %

    Ch

    an

    ge

    i

    n

    Tu

    mo

    r

    Bu

    rd

    en

    0 1 6 3 2 4 8 6 4

    - 1 0 0

    - 5 0

    0

    5 0

    1 0 0

    W e e k

    %

    Ch

    an

    ge

    i

    n

    Tu

    mo

    r

    Bu

    rd

    en

  • 0

    2 0

    4 0

    6 0

    8 0

    n=38

    2.1

    0

    2 0

    4 0

    6 0

    8 0

    % T

    umor

    cells

    Pos

    itive

    Benefit(n = 21) No Benefit(n=38)

    1.6% 3.6%

    PD-L1 testing was performed using the BMS developed PD-L1 IHC method (Dako PD-L1 IHC 28-8 pharmDx assay); PD-L1+ defined as ≥ 1% of tumor cells

    PD-L1 Expression of Baseline Biopsy Samples

    Patients with Ovarian CancerOnly • 34% had any tumor PD-L1 expressionSimilar PD• -L1 expression in Benefited and No Benefit groups

    Patient with CRC • Only 13% had any tumor PD-L1 expression

    Includes all patients with ovarian cancer and CRC

    This patient had 28% reduction in tumor, then PD

    9Rachel E. Sanborn

  • Generally Low TIL Levels in Baseline Biopsy SamplesOvarian Cohort (all pts) - Baseline CRC Cohort (all pts)- Baseline

    10Rachel E. Sanborn

    0

    2 0

    4 0

    6 0

    8 0 CD4

    11.7 9.1

    Benefit (n = 16) No Benefit (n=37)

    0

    2 0

    4 0

    6 0

    8 0

    CD8

    9.1 7.3

    Benefit (n = 21) No Benefit (n=38)

    0

    1 0

    2 0

    3 0

    FoxP3

    Benefit (n = 20) No Benefit (n=39)

    1.4 1.5

    0

    2 0

    4 0

    6 0

    8 0 CD4

    3.1

    (n=48)

    0

    1 0

    2 0

    3 0

    FoxP3

    2.7

    (n=39)

    0

    2 0

    4 0

    6 0

    8 0 CD8

    5.0

    (n=39)

  • Combination Treatment Increases Tumor PD-L1 and CD8 in Patients with Ovarian Cancer

    IHC analysis of all patients with ovarian cancer and CRC with paired biopsies (On• -study ̴day 29)Patients with ovarian cancer had significant increase in tumor expression of PD• -L1 and CD8+ TIL

    14 • of 23 patients (61%) had increase in PD-L1 and 14 of 24 patients (58%) had increase in CD8These changes were rarely observed in CRC patient samples•

    (dashed line represents mean)

    11Rachel E. Sanborn

    0

    10

    20

    30

    40

    50

    60

    70

    80

    1 20

    10

    20

    30

    40

    50

    60

    70

    80

    1 2Baseline On-study

    CRCOVARIAN

    % P

    ositi

    ve T

    umor

    Cel

    ls (n=24)(n=23)

    Baseline On-study0.38Paired T-Test P= 0.021

    PD-L

    1

    0

    10

    20

    30

    40

    50

    60

    1 20

    10

    20

    30

    40

    50

    60

    1 2Baseline On-study

    CRCOVARIAN

    % P

    ositi

    ve T

    umor

    Cel

    ls (n=24)(n=24)

    Baseline On-study0.39Paired T-Test P= 0.012

    CD8

  • Enhanced PD-L1 Expression and CD8 TIL is Associated with Better Outcome in Ovarian Patients

    12Rachel E. Sanborn

    All ovarian patients with paired biopsy samples (Day 29)Benefit/No Benefit PFS

    Median (mos) PR+uPR7.4 38%3.5 13%

    HR 0.32p-value 0.066

    Median (mos) PR+uPR7.4 42%2.6 0%

    HR 0.24p-value 0.015

    0

    10

    20

    30

    40

    50

    60

    70

    1 20

    10

    20

    30

    40

    50

    60

    70

    1 2

    % P

    ositi

    ve T

    umor

    Cel

    lsBenefit (n=14) No Benefit (n=9)

    Baseline On studyBaseline On study

    PD-L

    1

    0.016 0.52Paired T-Test P=

    PR

    PR

    PR

    PR

    0

    10

    20

    30

    40

    50

    60

    1 20

    10

    20

    30

    40

    50

    60

    1 2

    % P

    ositi

    ve T

    umor

    Cel

    ls

    Benefit (n=14) No Benefit (n= 10)

    Baseline On studyBaseline On study

    CD8

    0.220.025Paired T-Test P=

    PR

    PR

    PRPR

    (dashed line represents mean)

    0 5 1 0 1 5 2 0

    0

    5 0

    1 0 0

    M o n t h s

    P D - L 1 i n c r e a s e < 5 % ( n = 1 5 )

    P D - L 1 i n c r e a s e > = 5 % ( n = 8 )

    %

    Pr

    og

    re

    ss

    io

    n-

    fr

    ee

    s

    ur

    vi

    va

    l

    0 5 1 0 1 5 2 0

    0

    5 0

    1 0 0

    M o n t h s

    C D 8 i n c r e a s e < 5 % ( n = 1 2 )

    C D 8 i n c r e a s e > = 5 % ( n = 1 2 )

    %

    Pr

    og

    re

    ss

    io

    n-

    fr

    ee

    s

    ur

    vi

    va

    l

  • 13

    0

    20

    40

    60

    80

    100

    0

    10

    20

    30

    40

    Baseline On study

    Baseline On study

    Perc

    ent T

    IM3+

    Celll

    sPe

    rcen

    t LAG

    -3+

    Celll

    s

    Expression of exhaustion markers

    Expression of TIM-3, LAG-3 Does Not Distinguish Refractory Ovarian Patients

    100

    50

    -100

    -50

    0

    0 10 20 30 40 50 Weeks

    Targ

    et T

    umor

    Size

    % C

    hang

    e Fr

    om B

    asel

    ine PD

    SDPR

    Biopsy0

    10

    20

    30

    40

    50

    60

    Perc

    ent C

    D8+

    Celll

    s

    Analysis on ovarian patients with CD8 TIL but varying clinical outcome

    Baseline On study

    The inhibitory/exhaustion markers TIM-3 and LAG-3 increased in correlation with CD8 T cells independent of outcome

    Rachel E. Sanborn

  • Analysis of peripheral blood

    T cells and NK cells had marked up• -regulation of HLA-DR

    Stark • Treg decreases were seen in majority of patients

    Rapid increase in chemokines (CCL• 2, CCL4, CXCL9)

    Changes observed across all varlilumab dose cohorts and •similar between patients with CRC and Ovarian cancer

    No apparent correlation of peripheral blood parameters that •we investigated with clinical outcome

    14Rachel E. Sanborn

    -100

    0

    100

    200

    300

    Pre Dose Day 2

    Seru

    m C

    XCL9

    % Δ

    Fro

    m B

    asel

    ine

    Baseline Day 14

    Treg

    % Δ

    From

    Bas

    elin

    e

    -100.0

    -50.0

    0.0

    50.0

    100.0

    -50

    0

    50

    100

    150

    C1DS1 C1DS2

    CD4

    HLA

    -DR+

    %

    ΔFr

    om B

    asel

    ine

    Baseline Day 14

    Baseline Day 2

  • ConclusionsVarlilumab and nivolumab combination therapy was generally well tolerated at all varlilumab •dose levels testedPatients’ baseline tumor biopsies were mostly “cold” (PD• -L1 neg. or low and low TIL) with low expectation of responding to check-point inhibition monotherapy Uniquely in ovarian cancer cohort, increased PD• -L1 and CD8 TIL were observed in ̴ 60% of patients with paired biopsy samples

    The increase in PD− -L1 and CD8 TIL is associated with better clinical outcome3 • mg/kg dosing of varlilumab may have more clinical activity than other doses studied. Among CRC patients, durable clinical responses were observed in a patient with MSI• -High tumor and one with a high mutational burdenOpportunities for further evaluation of varlilumab/nivolumab:•

    Molecular analysis may identify biomarkers in baseline biopsies of ovarian patients whose −tumors can be predicted to change from “cold” to “hot”Expanding the experience in CRC (or other tumors) with MSI− -High or high mutational burden

    15Rachel E. Sanborn

  • Acknowledgments

    • Study CDX1127-02 investigators and their staff

    Thank you to the patients and their families who participated in this study

    16Rachel E. Sanborn

    Rachel SanbornMichael PishvaianMargaret CallahanAmy WeiseBranimir I SikicOsama Rahma

    Daniel ChoNaiyer RizviJose LutzkyRhonda L. BittingAlexander StarodubAntonio Jimeno

    • Celldex Therapeutics

    • Bristol-Myers Squibb

    Anti-CD27 Agonist Antibody Varlilumab with Nivolumab �for Colorectal and Ovarian Cancer: �Phase 1/2 Clinical Trial Results�Combination of CD27 Costimulation with PD-(L)1 BlockadeSlide Number 3Baseline Patient CharacteristicsVarlilumab & Nivolumab Combination Therapy is Well ToleratedTumor Response: Ovarian CancerTumor Response: CRCAnalysis of Immune Monitoring and Correlations with Outcome PD-L1 Expression of Baseline Biopsy SamplesGenerally Low TIL Levels in Baseline Biopsy SamplesCombination Treatment Increases Tumor PD-L1 and CD8 in Patients with Ovarian CancerEnhanced PD-L1 Expression and CD8 TIL is Associated with Better Outcome in Ovarian PatientsSlide Number 13Analysis of peripheral bloodConclusionsAcknowledgments