Download ReACT: Long-term survival from a randomized phase II study of

ReACT: Long-term survival from
a randomized phase II study of
rindopepimut (CDX-110) plus
bevacizumab in relapsed
glioblastoma
David A. Reardon1, Annick Desjardins2, James Schuster3, David D. Tran4, Karen L. Fink5,
Louis B. Nabors6, Gordon Li7, Daniela A. Bota8, Rimas V. Lukas9, Lynn S. Ashby10, J. Paul Duic11,
Maciej M. Mrugala12, Andrea Werner13, Laura Vitale13, Yi He13, Jennifer Green13, Michael J. Yellin13,
Christopher D. Turner13, Thomas A. Davis13, John H. Sampson2, The ReACT Study Team
1. Dana‐Farber Cancer Institute, Boston, MA; 2. Duke University Medical Center, Durham, NC; 3. University of Pennsylvania, Philadelphia, PA; 4. Washington University, St. Louis, MO; 5. Baylor Research Institute, Dallas, TX; 6. University of Alabama at Birmingham, Birmingham, AL; 7. Stanford University School of Medicine, Stanford, CA; 8. UC Irvine Medical Center, Irvine, CA; 9. University of Chicago, Chicago, IL; 10. Barrow Neurological Institute, Phoenix, AZ; 11. Long Island Brain Tumor Center at Neurological Surgery, P.C., Lake Success, NY; 12. University of Washington School of Medicine, Seattle, WA; 13. Celldex Therapeutics, Inc., Hampton, NJ
1
EGFR Mutation Variant III (EGFRvIII)
● Tumor-specific oncogene expressed in one-third of primary GBM, seldom
expressed with IDH mutations and not found in normal tissue
● EGFRvIII(+) cells may induce growth in EGFRvIII(-) cells via paracrine
signaling, membrane-derived microvesicles, and tumor stem cells1-4
EGFRvIII Linked To Poor Long Term Survival
Retrospective analysis of newly diagnosed glioblastoma patients treated with
standard temozolomide in the RTOG 0525 trial (data provided by NRG Oncology)
Overall Survival (%)
100
Median OS, months
(from randomization)
75
EGFRvIII- (n=147)
18.2
EGFRvIII+ (n=62)
14.3
HR = 1.68 (1.23, 2.29)
p < 0.001
50
Median time from
recurrence to death for
patients with EGFRvIII(+)
tumors was 8.7 months
(95% CI: 7.6-10.3).
25
0
1.
2.
3.
4.
Inda, Genes Dev. 2010
Al-Nedawi, Nat Cell Biol. 2008
Wong, JCO 2008
Fan, Cancer Cell 2013
0
12
24
36
48
60
72
84
96
Months from Study Randomization
108 120
2
Rindopepimut
Rindopepimut consists of
EGFRvIII peptide
conjugated to Keyhole
Limpet Hemocyanin (KLH)
Proposed Mechanism of Action
● Generates a
specific immune
response against
EGFRvIII-expressing
GBM
● “Ready to use”
formulation
● Delivered as
intradermal injection
of 500 µg
rindopepimut with
150 µg GM-CSF
as an adjuvant
EGFRvIII‐specific cytotoxic T lymphocytes (CTLs) may have a role in tumor destruction with rindopepimut, but it has been difficult to clearly demonstrate the presence of these cells.
3
Rationale for Rindopepimut Plus
Bevacizumab in Relapsed GBM
● Promising PFS/OS from Phase 2 studies in newly diagnosed,
resected, EGFRvIII-expressing GBM1-3
● Anecdotal evidence suggests that rindopepimut may induce specific
immune responses and regression in multifocal and bulky tumors
– Marked tumor regression with rindopepimut in combination with standard
treatments (compassionate use experience)
● Bevacizumab (BV) may optimize EGFRvIII-specific immune
response4-6
– VEGF may mediate
immunosuppression (impairs DC
maturation, alters tumor
endothelium, potentially decreasing
immune cell infiltration)
Expected Outcome for Relapsed
GBM Treated with BV7
ORR
(%)
28
PFS6 Median PFS Median OS
(%)
(months)
(months)
43
4.2
9.2
– BV enhances immune-mediated
anti-tumor effect in tumor models
1.
2.
3.
4.
Sampson, JCO 2010
Sampson, Neuro-Onc 2010
Schuster, Neuro-Onc 2015
Johnson, Expert Opin. Biol. Ther. 2007
5. Shrimali, Cancer Research 2010
6. Osada, Cancer Immunol Immunother 2008
7. Friedman, JCO 2009
4
Study Design
BV-naïve EGFRvIII+ GBM,
1st or 2nd relapse (n=70)
No prior BV or VEGF/VEGFR therapy
Randomization (1:1)
Double-blind treatment
Rindopepimut + BV
Control (KLH) + BV
Study Days
1 15 29 43 57 71 85 99 113
Disease assessment with MRI (every 8 weeks)
Bevacizumab (10 mg/kg IV every 2 weeks)
Study vaccine (500 µg rindopepimut with
150 µg GM-CSF) or
Control (100 µg KLH )
1. Friedman, JCO 2009
2. Wen, JCO 2010
Additional eligibility requirements
● EGFRvIII per centralized IDE-approved RT-PCR
assay
● Prior conventional radiation and TMZ
● ≤ 4 mg of dexamethasone daily
● No gliomatosis cerebri, infratentorial,
leptomeningeal or metastatic disease
● No prior intracerebral agents, antibody-based
therapy within 28 days, non-protein based agents
within 14 days, or radiation within 3 months of entry
Study Analyses
● Primary Analysis: PFS at 6 mos (PFS6) for intentto-treat (ITT) population
‒ Study Design: PFS6 of 40%1 vs 60%, 1-sided
 = 0.2, power = 80%
‒ Assessed by blinded expert review committee
● Secondary Analyses: ORR, PFS, OS, safety and
tolerability, EGFRvIII-specific immune response
● Tumor response evaluation by RANO criteria:2
assessment incorporates radiographic data, steroid
use and clinical status
● . Patients enrolled between May 2012 and May 2014
● Cut-off date for this long-term update: Sept 1, 2015
5
Patient Characteristics
Rindopepimut +
BV (n=36)
Control + BV
(n=37)
Age, years (median [range])
≥50 years (n [%])
59 (44-79)
35 (97%)
55 (30-75)
27 (73%)
Male (n [%])
19 (53%)
22 (59%)
2 (6%)
13 (36%)
14 (39%)
7 (19%)
5 (14%)
13 (35%)
12 (32%)
7 (19%)
35 (97%)
35 (95%)
10.8 (3.7-55.2)
11.6 (4.7-38.3)
33 (92%)
3 (8%)
28 (76%)
9 (24%)
Surgery after last relapse (n [%])
Gross-total resection
Partial resection/unspecified
15 (42%)
14 (39%)
1 (3%)
10 (27%)
6 (16%)
4 (11%)
On steroids at study entry (n [%])
18 (50%)
19 (51%)
KPS (n [%])
100
90
80
70
Primary GBM (n [%])
Time from diagnosis to study entry,
months (median [range])
Relapses (n [%])
1
2
6
Safety
Most Frequent Adverse Events
● Mean (range) number of
vaccinations:
‒ Rindopepimut + BV: 9.9 (3, 42)
‒ Control + BV: 6.5 (2, 24)
● Rindopepimut + BV was welltolerated
‒ No unexpected toxicity
associated with BV
administration
‒ No SAEs attributed to
rindopepimut
‒ No discontinuations due to
rindopepimut treatment-related
AEs
‒ Frequent grade 1-2 injection
site reactions (primarily
erythema and pruritus)
‒ One G2 hypersensitivity
reaction
‒ No evidence of increased
cerebral edema
(Regardless of relationship to study treatment)
Rindopepimut + BV
(n=35)
Control + BV
(n=37)
≥Grade 3 Overall ≥Grade 3 Overall
Arthralgia
-
8(23%)
1(3%)
2(5%)
Back pain
2(6%)
6(17%)
-
3(8%)
Convulsion
4(11%)
8(23%)
-
9(24%)
6(17%)
-
2(5%)
Diarrhea
Fall
Fatigue
Headache
1(3%)
1(3%)
6(17%)
1(3%)
2(5%)
9(26%)
2(5%)
10(27%)
8(23%)
2(5%)
9(24%)
Hemiparesis
-
2(6%)
2(5%)
6(16%)
Hyperglycemia
-
3(9%)
3(8%)
4(11%)
8(23%)
3(8%)
9(24%)
1(3%)
4(11%)
Hypertension
1(3%)
Nausea
-
8(23%)
Vomiting
-
6(17%)
-
2(5%)
Includes any adverse event occurring at ≥15% frequency, or in >2 patients at
severity Grade ≥3, in either treatment group (excluding injection site reactions).
7
Progression-Free Survival
PFS6 (Crude rate)
Primary Endpoint:
Expert Review
Rindopepimut + BV
10/36 (28%)
Control + BV
6/37 (16%)
Investigator
Assessment
10/36 (28%)
p = 0.1163*
p = 0.0658*
5/37 (14%)
* Chi-square test (1-sided). Study is designed to detect a PFS6 difference with 1-sided  = 0.2.
Progression-Free Survival (Kaplan-Meier Method)
Expert Review
75
HR = 0.72 (0.43, 1.21)
p = 0.2187**
50
25
0
0
6
12
18
24
Investigator Assessment
100
Progression-Free Survival (%)
Progression-Free Survival (%)
100
30
Months
Per-protocol population analysis:
HR = 0.61 (0.35, 1.05)
p = 0.0733
36
42
Rindopepimut + BV
Control + BV
75
HR = 0.62 (0.38, 1.03)
p = 0.0654**
50
25
** Logrank test (2-sided)
0
0
6
12
18
24
30
36
42
Months
Per-protocol population analysis:
HR = 0.53 (0.31, 0.91)
p = 0.0186
8
Radiographic Response
Expert Review
Rindopepimut
+ BV
Control
+ BV
Rindopepimut
+ BV
Control
+ BV
9/30 (30%)
6/34 (18%)
7/31 (23%)
3/32 (9%)
11/30 (37%)
8/34 (24%)
11/31 (36%)
9/32 (28%)
ORR (confirmed CR/PR)
Any response (≥50% shrinkage)
including those not sustained at
subsequent assessment
Duration of Response (months [95% CI])
Investigator Assessment
7.8 (3.5, 22.2) 5.6 (3.7, 7.4)
9.0 (3.7, 20.5) 7.4 (3.7, 17.4)
Response-evaluable patient subset with measurable disease.
Duration of Response
Rindopepimut
+ BV
Expert Review (Confirmed Responses)
CR
PR
CR
PR
Control
+ BV
*
Censored response
0
6
12
Months
18
24
* Excluded from PP population
9
Reduction in Steroid Use
Rindopepimut +
BV
Control +
BV
Able to stop steroids for ≥2 months*
9/18 (50%)
5/19 (26%)
Able to stop steroids for ≥6 months*
6/18 (33%)
0 (0%)
*Subset on steroids at study entry
Duration off Steroids
36
Months
30
Patients remain off steroids
24
18
12
6
0
Rindopepimut + BV
Control + BV
10
Overall Survival
Rindopepimut + BV
Control + BV
Median, months
(95% CI)
OS 12
OS 18
OS 24
11.3 (9.9, 16.2)
44%
32%
25%
9.3 (7.1, 11.4)
32%
13%
0%
HR = 0.53 (0.32, 0.88)
p = 0.0137*
Per-protocol population analyses:
HR = 0.53 (0.31, 0.90)
p = 0.0177*
Five patients in the rindopepimut + BV arm, and 1 patient in the control + BV arm, continue survival follow-up
without progression per expert review.
* Log-rank test (2-sided)
11
ReACT Results are Consistent with
Immunotherapy Experience
Overall Survival
100
Rindopepimut + BV
Control + BV
Overall Survival (%)
ReACT Data
Progression-Free Survival (%)
Progression-Free Survival
Rindopepimut + BV
75
Control + BV
50
25
0
0
6
12
18
24
30
36
42
Months
FDA approval
data –
ipilimumab
for metastatic
melanoma
Ipilimumab
Hodi NEJM 2010
Ipilimumab
Control
Control
12
Overall Survival: Sub-group Analysis
No surgery after last relapse
Overall Survival (%)
HR = 0.44 (0.23, 0.86)
P = 0.014
Rindopepimut + BV
Control + BV
All patients, by whether surgery after last relapse
100
HR = 1.19 (0.69, 2.06)
P = 0.54
75
Surgery
No surgery
50
Prior surgery/extent of
surgery did not portend a
better outcome in this
population.
25
0
0
6
12
18
24
30
36
42
Months
Favors rindopepimut
Favors control
For patients surviving ≥ 18 months on study: 6/9 (67%) in the
rindopepimut+BV arm vs. 1/4 (25%) in the control+BV arm had
last relapse-free interval prior to study entry of ≤ 6 months.
Overall Survival (%)
One prior relapse
HR = 0.55 (0.32, 0.96)
P = 0.031
Rindopepimut + BV
Control + BV
13
Rindopepimut Elicits Potent Anti-EGFRvIII Antibody
Response that is Associated with Prolonged Survival
and Mediates ADCC
Early (Week 8) Antibody Response
Titer ≥1:12,800 by Week 8
All others
75
HR (95% CI): 0.53 (0.24, 1.16)
p = 0.1065
50
25
0
0
6
12
18
24
30
36
100
Overall Survival (%)
Overall Survival (%)
100
Overall Antibody Response
75
HR (95% CI): 0.17 (0.07, 0.45)
p < 0.0001
50
25
0
0
42
Titer ≥1:12,800 at any time
All others
6
Months
12
18
24
30
36
42
Months
Anti-EGFRvIII Antibody Titers
ADCC Activity Against EGFRvIII+ Tumor Cells
Limit of Quantification
% Killing of EGFRvIII+ Cells
Geometric Mean Titer (1:X)
60
50
40
30
20
10
0
1:50
Weeks
1:100 1:200 1:400 1:800 1:1600 1:3200
Patient serum dilution (from peak Ab response)
14
Conclusions
● Mature randomized survival data continue to show a marked benefit
(HR=0.53, p=0.0137) with long-term survival in this small trial
‒ Long term survival not predicted by prognostic patient
characteristics
‒ Advantage to rindopepimut therapy across multiple endpoints
including long-term progression-free survival, objective response
rate and steroid requirement
● The addition of rindopepimut to bevacizumab did not significantly
increase toxicity
‒ Associated with reduction in steroid usage
● Remarkable frequency and level of anti-EGFRvIII immune responses
despite prior chemotherapy and growing tumor
‒ EGFRvIII antibody response correlates with potent effector
function and improved clinical outcome
● Activity profile consistent with prior immunotherapy experience1
● Phase 3 trial (ACT IV) in newly diagnosed completed accrual Dec 2014
‒ Interim analysis expected in early 2016, with timing of final data
dependent on events
1. Hodi, NEJM 2010
15
Acknowledgements
The ReACT patients and their families
The ReACT Study Investigators and Coordinators
Ron Steis & Kathy Frank
Lynn Ashby & Andrea Ralph
Karen Fink & Sara Gonzales
David Peereboom & Marci Ciolfi
David Reardon & Karly Griffin
Laszlo Mechtler & Olga Ananina
Annick Desjardins & Shelly Dutton
Samuel Goldlust & Lori Cappello
John Trusheim & Meghan Hultman
Michael Lim & Susan Zhen
Richard Green & Sandra Baker‐Bolden
Tara Morrison & Yasmin Candelo
J. Paul Duic & R. Kimberly Prabhu
Jorg Dietrich & Tucker Cushing
Atif Hussein & Norma Cuervo
Joseph Landolfi & Charles Probeni
Ryan Merrell & Nancy Miedzianowski
Nicholas Avgeropoulos & Mollie Geismer
Jeff Chen & Susan Staat
Erin Dunbar & Michelle Humphreys
Heinrich Elinzano & Wendy Turchetti
Nina Paleologos & Amanda Begley
Gordon Srkalovic & Vicki Gilreath
Gordon Li & Sophie Bertrand
Agnieszka Kowalska & Susan Fiore
Tara Benkers & Nathan Hansen
Lawrence Berk &Sharon Porth
Morris Groves & Marlena Griffin
Nicholas Butowski & Thelma Munoz
Andrew Sloan & Gina Cascone
Louis Nabors & Thiru Pillay
Daniela Bota & Jinah Chung
Independent Expert Radiographic Review Panel Raymond Y. Huang, MD, PhD
Associate Radiologist, Dept. of Radiology/Neuroradiology Brigham and Women's Hospital
Whitney Pope, MD, PhD
Associate Professor, Radiology
Director, UCLA Brain Tumor Imaging
David Geffen School of Medicine at UCLA
John deGroot, MD (adjudication)
Associate Professor, Department of Neuro‐Oncology
Director, Neuro‐Oncology Fellowship Program Director, Clinical Research, Department of Neuro‐Oncology
University of Texas MD Anderson Cancer Center
Maciej Lesniak, Rimas Lukas, Kelly Nicholas & Christina Amidei
Richard Curry & Suzanne Sifri
Denise Damek & Monica Robischon
Larry Junck & Adam Booth
James Schuster, Donald O’Rourke & Chau Nguyen
Jan Drappatz & Melinda Vargas‐Jaffe
Nimish Mohile & Jennifer Serventi
Sigmund Hsu & Greg Lu
Maciej Mrugala & Ashley Waldie
Naveed Wagle & Aida Lozada
Nitin Chandramouli & Robbyn Oliver
Reid Thompson & Tom Leonard‐Martin
Glenn Lesser & Jennifer MacLean
David Tran & Madelyn Kissel
Celldex Therapeutics, Inc. The ReACT study team
For his pioneering work and continued collaboration in the development of rindopepimut: John H. Sampson, MD, PhD
Professor of Neurosurgery and Chief of the Division of Neurosurgery at Duke University Medical Center
16