Upload
owen-ford
View
216
Download
0
Tags:
Embed Size (px)
Citation preview
Putting Cancer in Check with Immunotherapy: Melanoma and Beyond
Michael Postow, MDMelanoma and Immunotherapeutics Service
Memorial Sloan Kettering Cancer Center
Bristol-Myers Squibb: -Research support-Participated in an advisory council (non-paid)
Amgen:-Participated in an advisory council (non-paid)
2013 Breakthrough of the Year
What happened in 1891?
What happened in 1891?
Main Questions
1. How can the immune system treat cancer?
2. What have we learned from clinical experience in melanoma?
3. How can we improve?
Mellman et al. Nature 2011
Immunity in tumor control
Mellman et al. Nature 2011
Immunity in tumor control
Vaccines
Cytokines
Adoptive cell transfer
Immunomodulatory antibodies
Median OS benefit: 4.1 monthsHR : 0.78 (95% CI: 0.61-0.98; P = .03)
Sipuleucel-T vaccine improves survival in metastatic prostate cancer
Kantoff PW, et al. N Engl J Med. 2010
Prob
abili
ty o
f Sur
viva
l (%
)
Mos Since Randomization
80
60
40
20
0
100
0 12 24 36 48 60 72
Sipuleucel-T (n = 341)Median Survival: 25.8 Mos.
Placebo (n = 171)Median Survival: 21.7 Mos.
HD IL-2 Therapy: Durable Responses
Atkins MB, et al. J Clin Oncol. 1999
Metastatic Melanoma (N = 270) Metastatic RCC (N = 255)
-HD IL-2 produces durable responses in 6% to 10% of patients with advanced melanoma or RCC-Few relapses in patients responding for over 2.5 years (likely cured)-FDA approval in 1992 (RCC) and 1997 (melanoma)
1.0
0.8
0.6
0.4
0.2
0.0Prob
abili
ty o
f Con
tinui
ng R
espo
nse
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Duration of Response (Mos)
CR (n = 17)PR (n = 26)CR + PR (n = 43)
1.0
0.8
0.6
0.4
0.2
0.0Prob
abili
ty o
f Con
tinui
ng R
espo
nse
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Duration of Response (Mos)
CRPR All
140 150 160 170 180
Mellman et al. Nature 2011
Immunity in tumor control
Vaccines
Cytokines
Adoptive cell transfer
Immunomodulatory antibodies
Turning up The Activating Blocking the Inhibiting
Ways to keeping the T cells “active”
Mellman et al. Nature 2011
ARSS Question 1Which of the previously shown T cell co-regulatory targets has been targeted with FDA
approved antibodies?
1. CTLA-4
2. LAG-3
3. PD-1
4. CD 28
5. 1 and 3
6. 1, 3, and 4
7. 1, 2, 3, and 4
Turning up The Activating Blocking the Inhibiting
Ways to keeping the T cells “active”
Mellman et al. Nature 2011
DendriticCell T Cell
T-Cell Receptor
CD28
CTLA-4
B7
B7
Cytotoxic T Lymphocyte Antigen 4
MHC with Antigen
Lymph Node
Postow et al. JCO 2015
DendriticCell T Cell
PD-L1/PD-L2
PD-1 Immune Checkpoints
TumorCD28B7
MHC and Antigen T-Cell Receptor
PD-1
PD-L1
PD-L1
PD-1 PD-L1
Lymph Node Tumor Microenvironment
PD-1
Postow et al. JCO 2015
Blocking immunologic checkpoints
NivolumabPembrolizumab
PidilizumabIpilimumab and Tremelimumab
Kyi and Postow FEBS Letters 2014
MPDL3280AMEDI4736
MSB0010718C
(1) Hodi et al. NEJM 2010(2) Hamid et al. NEJM 2013(3) Topalian et al. NEJM 2012(4) Weber et al. Lancet Oncol 2015(5) Kelly et al. JCO 2001
Adverse Events
• CTLA-4: Rash, diarrhea, hepatitis, endocrine– 24% grade 3/4 (1)
• PD-1/PD-L1: Rash, fatigue, arthralgias– 6-12% grade 3/4 (2-4)
• Chemotherapy: Alopecia, nausea, myelosuppression – ~50% grade 3/4 (5)
Ipilimumab Rashes
Can be treated with topical corticosteroids
Diarrhea and Colitis
Slangen et al., World J Gastrointest Pharmacol Ther, 2013
Focal Active Colitis
Alterations in Crypt Epithelium
Maker AV, Ann Surg Oncol. 2005
Diarrhea and Colitis
Weber et al. JCO 2012, reprinted from Blansfield J Immunother 2005
Enlarged pituitary due to hypophysitis
Pneumonitis
3/30/20112/21/2011
Two doses of ipilimumab and four of nivolumab
ARSS Question 2A 55 year old man has completed 3 doses of ipilimumab and
has developed significant diarrhea. Which do you recommend?
1. Ciprofloxacin and Flagyl
2. A several week course of oral steroids
3. Observation for 1 week, and then oral steroids only if worsening to avoid blunting of immunotherapy effect.
4. Infliximab
Weber JS, J Clin Oncol 2012
Kinetics of irAEs with ipilumumab
Rash, pruritisLiver toxicityDiarrhea, colitisHypophysitis
Toxic
ity G
rad
e
Wks
14
0 2 4 6 8 10
12
Diarrhea/Colitis Management
IPILIMUMAB
IPI
IPI
IPI
Blocking immunologic checkpoints
NivolumabPembrolizumab
PidilizumabIpilimumab and Tremelimumab
Kyi and Postow FEBS Letters 2014
MPDL3280AMEDI4736
MSB0010718C
Antibodies that block CTLA-4
Tremelimumab
Ipilimumab
Median OS 10.1 mos
24% alive at 2 years
Hodi et al. NEJM 2010
Response rate of 10.9%
Ipilimumab confers OS benefit to gp100 vaccine in phase III study
Long-term results of 2nd phase III study
Maio et al. J Clin Oncol 2010
Median OS 11.2 vs. 9.1 mosAt 5 years, 18.2 vs. 8.8% alive, p=0.002
Ipilimumab “responses” can be delayed
10 mg/kg ipilimumab
Q3W X 4
Pre-treatment
Week 36: Still Regressing
Week 12: Progression
Week 20: Regression
New lesions
Wolchok ASCO 2008
July 2006
Immune-related response criteria
• New lesions are not counted automatically as progressive disease.
• Increase in tumor burden must be confirmed on subsequent scan.
Wolchok et al. Clin Cancer Res 2009
Blocking immunologic checkpoints
NivolumabPembrolizumab
PidilizumabIpilimumab and Tremelimumab
Kyi and Postow FEBS Letters 2014
MPDL3280AMEDI4736
MSB0010718C
DendriticCell T Cell
PD-L1/PD-L2
PD-1 Immune Checkpoints
TumorCD28B7
MHC and Antigen T-Cell Receptor
PD-1
PD-L1
PD-L1
PD-1 PD-L1
Lymph Node Tumor Microenvironment
PD-1
Postow et al. JCO 2015
PD-1/PD-L1 Agents in Development
Target Agent Class
PD-1
Nivolumab (MDX1106, BMS-936558)
IgG4 fully human Ab
Pembrolizumab (MK-3475)
IgG4 engineered humanized Ab
Pidilizumab (CT 011)‑ IgG1 humanized Ab
AMP 224‑ Fc of human IgG-PD-L2 fusion
PD-L1
BMS935559 (MDX 1105)‑ IgG4 fully human Ab
MPDL3280A IgG1 engineered fully human Ab
MEDI4736 IgG1 engineered fully human Ab
MSB0010718C IgG1 fully human Ab
ARSS Question 3Which of the following statements about anti-PD-1 therapy in
melanoma is false?
1. Nivolumab has improved overall survival compared to dacarbazine.
2. Pembrolizumab has improved overall survival compared to ipilimumab.
3. Pembrolizumab has a higher rate of significant toxicity than ipilimumab.
4. Nivolumab and pembrolizumab are only available for patients who have progressed on ipilimumab and if relevant, a BRAF inhibitor.
Responses with Nivolumab
Topalian et al, NEJM 2012Nivolumab: 1mg/kg every 2 weeks in melanoma patients
Response rate:~30%
Nivolumab improves OS
Nivolumab vs. DTIC-- HR: 0.42
Ipi + DTIC vs. DTIC-- HR: 0.72
Pembrolizumab: Clinical Activity Baseline: April 13, 2012
Images courtesy of A. Ribas, UCLA.
72-year-old male with symptomatic progression after bio-chemotherapy, HD IL-2, and ipilimumab
April 9, 2013
Pembrolizumab improves OS compared to ipilimumab
Drug Sponsor Target Disease Type Response (n) Reference
Nivolumab BMS PD-1
Solid Tumors 21% (42) Topalian et al. NEJM 2012
Melanoma 32% (44) Weber et al. JCO 2013
NSCLC 14% (63) Antonia et al. WCLC 2013
RCC 21% (168) Motzer et al. ASCO 2014
Ovarian 17% (18) Hamanishi ASCO 2014
Pembrolizumab Merck PD-1
Melanoma 40% (113) Daud et al. AACR 2014
NSCLC 19% (146) Gandhi et al. AACR 2014
Melanoma 34% (411) Ribas et al. ASCO 2014
NSCLC 26% (45) Rizvi et al. ASCO 2014
Head & Neck 18% (55) Selwert et al. ASCO 2014
CT-011 Curetech PD-1 Hematologic Cancers 33% (17) Berger et al. Clin Cancer Res 2008
Melanoma 6% (101) Atkins et al. ASCO 2014
AMP-224 Amplimmune/GSK
PD-1 Solid tumors Response, SD (42) Infante et al. ASCO 2013
Efficacy of PD-1 Agents
Drug Sponsor Target Disease Type Response (n) Reference
Nivolumab BMS PD-1
Solid Tumors 21% (42) Topalian et al. NEJM 2012
Melanoma 32% (44) Weber et al. JCO 2013
NSCLC 14% (63) Antonia et al. WCLC 2013
RCC 21% (168) Motzer et al. ASCO 2014
Ovarian 17% (18) Hamanishi et al. ASCO 2014
Pembrolizumab Merck PD-1
Melanoma 40% (113) Daud et al. AACR 2014
NSCLC 19% (146) Gandhi et al. AACR 2014
Melanoma 34% (411) Ribas ASCO 2014
NSCLC 26% (45) Rizvi et al. ASCO 2014
Head & Neck 18% (55) Selwert et al. ASCO 2014
CT-011 Curetech PD-1 Hematologic Cancers 33% (17) Berger et al Clin Cancer Res 2008
Melanoma 6% (101) Atkins ASCO 2014
AMP-224 Amplimmune/GSK
PD-1 Solid tumors Response, SD (42) Infante et al. ASCO 2013
Efficacy of PD-1 Agents
Drug Sponsor Target Disease Type Response (n) Reference
MPDL3280a Genentech PD-L1
Solid Tumors 21% (103) Herbst et al. ASCO 2013
Melanoma 23% (30) Hamid et al. ASCO 2013
NSCLC 23% (53) Sorial et al. ECC 2013
Bladder 26% (65) Powels et al. ASCO 2014
MEDI4736 MedImmune PD-L1
Solid Tumors 11% (179) Segal et al. ASCO 2014
NSCLC 16% (58) Brahmer et al ASCO 2014
Head & Neck 14% (22) Segal et al. ASCO 2014
Gastric 19% (16) Segal et al. ASCO 2014
MSB0010718C EMD Serono PD-L1 Solid tumors Response (27) Heery et al. ASCO 2014
MDX - 1105 BMS PD-L1 Solid Tumors 17% (135) Brahmer et al. NEJM 2012
Efficacy of PD-L1 Agents
Drug Sponsor Target Disease Type Response (n) Reference
MPDL3280a Genentech PD-L1
Solid Tumors 21% (103) Herbst et al. ASCO 2013
Melanoma 23% (30) Hamid et al. ASCO 2013
NSCLC 23% (53) Sorial et al. ECC 2013
Bladder 26% (65) Powels et al. ASCO 2014
MEDI4736 MedImmune PD-L1
Solid Tumors 11% (179) Segal et al. ASCO 2014
NSCLC 16% (58) Brahmer et al ASCO 2014
Head & Neck 14% (22) Segal et al. ASCO 2014
Gastric 19% (16) Segal et al. ASCO 2014
MSB0010718C EMD Serono PD-L1 Solid tumors Response (27) Heery et al. ASCO 2014
MDX - 1105 BMS PD-L1 Solid Tumors 17% (135) Brahmer et al. NEJM 2012
Efficacy of PD-L1 Agents
96 y.o. female– Progressed on previous cetuximab– HPV negative, PD-L1 positive– Treatment ongoing at 8 weeks
Segal et al., ASCO 2014
Baseline Day 28
Response in Patient with Head and Neck Cancer
MPDL3280A: Urothelial Bladder Cancer
• Median time to first response was 42 days (range, 38 to 85 days)• Median duration of response has not been reached
Powles T, et al. ASCO 2014
Pembrolizumab: AEs in > 5% of Patients
Adverse Event (N = 135) All Grades, n (%) Grades 3/4, n (%)
Any 107 (79.3) 17 (12.6)
Fatigue 41 (30.4) 2 (1.5)
Rash 28 (20.7) 3 (2.2)
Pruritus 28 (20.7) 1 (0.7)
Diarrhea 27 (20.0) 1 (0.7)
Myalgia 16 (11.9) 0
Headache 14 (10.4) 0
Increased AST 13 (9.6) 2 (1.5)
Asthenia 13 (9.6) 0
Nausea 13 (9.6) 0
Vitiligo 12 (8.9) 0
Hypothyroidism 11 (8.1) 1 (0.7)
Increased ALT 11 (8.1) 0
Cough 11 (8.1) 0
Pyrexia 10 (7.4) 0
Chills 9 (6.7) 0
Abdominal pain 7 (5.2) 1 (0.7)
Ribas A, et al. ASCO 2013
PD-L1 tumor cell membranestaining: 5%,
heterogeneous0 80%,
homogeneous3%, focal
PD-L1 Negative PD-L1 Positive
Slide courtesy of Margaret Callahan
Immunohistochemistry for PD-L1 expression
ARSS Question 4Which of the following statements about PD-L1 is
true?
1. PD-L1 is required to benefit from anti-PD1 therapy.
2. PD-L1 is required to benefit from ipilimumab.
3. PD-L1 is tested using similar methodology across clinical trials.
4. PD-L1 may reflect an immunologically active tumor
Nivolumab in NSCLC: PD-L1 is not associated with overall survival
PD-L1 +
PD-L1 -
Julie Brahmer ASCO 2014
PD-L1 ≠ EGFR, BRAF, HER2, ER, KRAS
• Many different assays measure PD-L1 differently
• PD-L1 negative tumors can still respond
• PD-L1 is a dynamic immunologic marker
• Heterogeneity exists within individual patients [1]
[1] Madore et al. Pigment Cell Melanoma Res 2014
Pretreatment immunologic biomarkers
Square peg into a round hole?
Immune Checkpoint CombinationsChemotherapy
Carboplatin/Paclitaxel (Lynch et al., J Clin Oncol 2012)
Anti-angiogenic agentsBevacizumab (Hodi et al., Cancer Immunol Res 2014)
Hormonal TherapyExemestane (Vonderheide et al., Clin Cancer Res 2010)Androgen deprivation (MDACC)
Targeted therapyVemurafenib (Ribas et al., NEJM 2013)
Other immunotherapyGM-CSF (Hodi et al., JAMA 2014)Nivolumab (Wolchok et al., NEJM 2013)
RadiotherapyPostow et al. NEJM 2013
Ipilimumab + Nivolumab
Therapy, % ORR
Ipilimumab 7
Nivolumab 28
Combination 42
Wolchok et al., NEJM 2013
First occurrence of new lesion
Cohort 2: 1 mg/kg nivolumab + 3 mg/kg ipilimumab
All patients in concurrent cohorts 250
200
150
100
50
0
-50
-100C
han
ge in
targ
et
lesio
ns
from
baselin
ePatients
300250200150100
50
-20
-40
-60
-80
-100
0
Chan
ge in
targ
et le
sion
s fr
om
base
line
(%)
0 10 20 30 40 50 60 70 80 90 100 110 120
Weeks since treatment initiation
• Proof-of-concept for the promise of immune checkpoint blocking antibodies
• Durable responses in melanoma and other cancers
• Biomarkers inform biology but not yet treatment selection
• Combinations may be better than single agent
Summary
• How should we target the next T cell co-regulatory receptors?
Future Questions
Mellman et al. Nature 2011
• How should we target the next T cell co-regulatory receptors?
• What endpoints should the FDA consider for regulatory approval?
Future Questions
• How should we target the next T cell co-regulatory receptors?
• What endpoints should the FDA consider for regulatory approval?
• Can targeting T-cells enhance other cancer therapeutics?
Future Questions
1891 to 2015
1891 to 2015
Postow M, et al. New Engl J Med 2012
Case reports of the abscopal effect