1. Advances in Immunotherapy
for Non-Small Cell Lung Cancer
Jarushka Naidoo, MB BCH
Assistant Professor of Oncology
Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University
Bloomberg-Kimmel Institute for Cancer Immunotherapy
FLASCO Annual Meeting
May 18, 2018
5. Specificity: virtually infinite antigen recognition
Adaptability: based on tumor genetic & epigenetic changes
Memory: durable responses even after drug discontinuation
Universality: potential anti-tumor effect regardless of tumor type
The Human Immune System
The Ultimate Anti-cancer Therapy?
Tumor Types with Objective Response to Anti-PD-1/PD-L1
Melanoma Non-small cell lung carcinoma Renal Cell carcinoma
Urothelial carcinoma Head and Neck carcinoma Merkel Cell carcinoma
MSI-high Colorectal carcinoma Biliary Tract carcinoma Ovarian carcinoma
Breast carcinoma Anal carcinoma Mesothelioma
Gastric adenocarcinoma Hogkins Lymphoma Hepatocellular carcinoma
Naidoo et al, Ann Transl Med 2016
6. Immunotherapy in NSCLC
Second-Line Therapy: Nivolumab vs. Docetaxel
272 patients:
135 (nivolumab); 137 (docetaxel)
Brahmer J, et al. N Engl J Med. 2015 Borghaei H, et al. N Engl J Med. 2015
582 patients:
292 (nivolumab); 290 (docetaxel)
7. PD-L1 Immunohistochemical assay
PD-L1 >50%
All patients (PD-L1 >1%)
1034 patients: 345 (pembro 2mg/kg), 346 (pembro 10mg/kg), 343 (docetaxel)
Garon EB, et al. N Engl J Med; Herbst RS, et al. Lancet. 2016
Immunotherapy in NSCLC
Second-line Therapy: PD-L1+ NSCLC, KEYNOTE studies
8. 8
OS= HR, 0.73
(95% CI, 0.60, 0.89)
P = 0.0015
Median 11.2 mo
(95% CI, 9.3, 12.6)
Median 15.6 mo
(95% CI, 13.3, 17.6)
Atezolizumab
Docetaxel
Median 7.7 mo
(95% CI, 6.3, 8.9)
Median 8.9 mo
(95% CI, 7.4, 12.8)
OS= HR, 0.73
(95% CI, 0.54, 0.98)
P = 0.0383
Non-squamous Squamous
Months Months
Anti-PD-L1 in Second-Line
Atezolizumab vs. Docetaxel: OS by histology
Rittmeyer A et al. Lancet. 2017;389:255-265
9. Gandara DR, et al. bTMB in POPLAR & OAK
Atezolizumab Blood TMB subgroups
Progression-FreeSurvival(%)
• Blood TMB ≥16 population accounted for 27% of biomarker evaluable group (N = 158)
• PFS benefit with atezolizumab was observed in the TMB ≥16 subgroup
• No prognostic effect was observed: patients with bTMB ≥16 did not have improved
PFS compared with patients with bTMB <16 in the docetaxel arm
Interaction P = 0.036
Progression-FreeSurvival(%)
Blood TMB ≥16 Blood TMB <16
Atezolizumab (N = 216)
Docetaxel (N = 209)
+ Censored
Atezolizumab (N = 77)
Docetaxel (N = 81)
+ Censored
Months Months
Progression-FreeSurvival(%)
Gandara DR et al. ESMO 2017
10. How to Evaluate for Blood-based TMB
Poplar and Oak Studies
• A 394 gene-based NGS assay
- Retrospective test of plasma samples (phase 2 POPLAR; phase 3 OAK study)
• 211/273 samples from POPLAR and 583/797 samples from OAK were evaluable
– The association between TMB and efficacy was analyzed and the cut-point of
TMB ≥16 was selected based on POPLAR
– Validated in OAK
bTMB Computational Methodology and Study Design
Blood collection,
plasma isolation &
cfDNA extraction
bTMB
Sequencing
POPLAR
(training)
OAK
(validation)
• All base substitutions
with ≥0.5% allele
frequency
• Remove germline
polymorphisms &
predicted driver
mutations
Gandara DR et al. ESMO 2017
11. Outline
• Brief Introduction to Immunotherapy in NSCLC
• First-line Immunotherapy in NSCLC
• Immunotherapy for Stage III NSCLC
• Future Directions:
- Neoadjuvant Studies
- Translational Studies
12. Progression-Free Survival
Events,
n
Median,
months
HR
(95% CI)
P
Pembro 73 10.3 0.50
(0.37-0.68)
<.001
Chemo 116 6.0
62%
50%
0 3 6 9 12 15 18
0
10
20
30
40
50
60
70
80
90
100
Time, months
PFS,%
No. at risk
154 104 89 44 22 3 1
151 99 70 18 9 1 0
48%
15%
ST v1.1 by blinded, independent central review
, 2016
Ann Oncol. 2016;27(suppl 6): Abstract LBA8. Reck M, et al. N Engl J Med. 9 Oct 2016. [Epub ahead of print].
Overall Survival
80%
72%
0 3 6 9 12 15 18 21
0
10
20
30
40
50
60
70
80
90
100
Time, monthsOS,%No. at risk
154 136 121 82 39 11 0
151 123 106 64 34 7 0
2
1
70%
54%
Events,
n
Median,
months
HR
(95% CI)
P
Pembro 44 NR 0.60
(0.41-0.89)
.005
Chemo 64 NR
DMC recommended stopping the trial because of
superior efficacy observed with pembrolizumab
Data cut-off: May 9, 2016
Reck M, et al. Ann Oncol. 2016;27(suppl 6): Abstract LBA8. Reck M, et al. N Engl J Med. 9 Oct 2016. [Epub ahead of print
Progression-Free Survival
5.7 month PFS benefit
HR= 0.5, p<0.001
48% vs. 15% progression-free at 1 year
Median OS results not mature
HR= 0.6, p=0.005
70% vs. 54% alive at 1 year
Reck et al, N Engl J Med 2016
First-line Pembrolizumab in NSCLC
KEYNOTE 024: PFS and OS
Overall Survival
13. Pembro
Responders
n = 69
Chemo
Responders
n = 42
TTR, mo
median
(range)
2.2
(1.4-8.2)
2.2
(1.8-12.2)
DOR,
mo
median
(range)
NR
(1.9+ to
14.5+)
6.3
(2.1+ to
12.6+)
First-line Pembrolizumab in NSCLC
KEYNOTE 024: ORR
Reck et al, N Engl J Med 2016
Objective Response Rate Duration of Response
14. Ph III First-Line Pembrolizumab
KEYNOTE 042: Study Design
Data to be presented
ASCO 2018
16. TMB as a Biomarker
First-Line Nivolumab in NSCLC: PFS
High TMB Low/Intermediate TMB
Garon et al, ASCO 2017
17. First-line PD-1 Monotherapy in NSCLC
Summary
• Pembrolizumab
- Improved survival compared to chemotherapy in advanced
NSCLC patients
- Relevance: PD-L1 > 50% and in patients PD-L1 > 1%
• Nivolumab
- Did not improve survival compared to chemotherapy for
advanced NSCLC patients whose PD-L1 > 5%
18. Pembrolizumab 200 mg
Q3W for 2 years
+
Pemetrexed 500 mg/m2
+ Carboplatin AUC
5 mg/mL/min Q3W for 4
cycles
Pemetrexed 500 mg/m2
+ Carboplatin AUC
5 mg/mL/min
Q3W for 4 cycles
R
(1:1)a
N = 123
First-line Pembrolizumab+Chemo in NSCLC
KEYNOTE-021 Cohort G: Study Design
Pemetrexed
500 mg/m2
Q3W permitted
as
maintenance
therapy
Study Population
•Untreated stage IIIB or IV
nonsquamous NSCLC
•No activating EGFR mutation or
ALK translocation
•Provision of a sample for PD-L1
assessmenta
•ECOG PS 0 or 1
•No untreated brain metastases
•No ILD or pneumonitis requiring
systemic steroids
End Points
•Primary: ORR (RECIST v1.1 per blinded, independent central review)
•Key secondary: PFS: OS, safety, relationship between antitumor activity and PD-L1 TPS
Langer et al, Lancet Oncol 2017
19. 0 3 6 9 12 15 18 21 24 27
0
10
20
30
40
50
60
70
80
90
100
Time, months
OverallSurvival,%
60 57 55 51 46 44 36 22 7 1
63 58 57 51 43 39 29 18 9 0
No. at risk
HR= 0.59
77% vs. 69% alive at 1 year
Pembro/Chemo vs, Chemo
p=0.03
HR= 0.54
Pembro/Chemo vs, Chemo
57% vs. 35% PFS at 1 year
p=0.006
Progression-Free Survival Overall Survival
First-line Pembrolizumab+Chemo in NSCLC
KEYNOTE-021 Cohort G: PFS and OS
Langer et al, Lancet Oncol 2017
20. Δ24.8%
(95% CI, 7.2% 40.9%)‒
P = 0.0029a
Pembro + PC
Responders
n = 34
PC Alone
Responders
N = 20
Median (range)
duration of response,
mo
NR
(1.4+ to
22.7+)
NR
(2.8 to 23.7+)
Ongoing response, % 50 40
First-line Pembrolizumab+Chemo in NSCLC
KEYNOTE 021-G: ORR
Langer et al, Lancet Oncol 2017
22. Overall SurvivalProgression-Free Survival
First-line Pembrolizumab+Chemo in NSCLC
KEYNOTE 189: Ph III: PFS and OS
HR= 0.49
69% vs. 49% OS at 1 year
Pembro/Chemo vs. Chemo
p<0.001
HR= 0.52
34% vs. 17% PFS at 1 year
Pembro/Chemo vs. Chemo
p<0.001
Gandhi et al, NEJM 2018
27. • In advanced non-squamous NSCLC
- Carboplatin+Pemetrexed+Pembro is superior to
chemotherapy alone, independent of PD-L1
- Carboplatin+Paclitaxel+Atezo+Bevavizumab is superior to
chemotherapy + bevacizumab, independent of PD-L1
- Carboplatin+Paclitaxel+Atezo+Bevavizumab benefitted
patients with EGFR/ALK alterations
• PD-L1 > 50% is relevant to select
- Pembrolizumab alone vs. Pembrolizumab + Chemotherapy
First-line PD-1+Chemotherapy in NSCLC
Summary
34. • Durvalumab-Tremelimumab
- Was no better than chemotherapy in treatment-naïve
patients (press release)
• Ipilumumab-Nivolumab
- Demonstrates superior PFS vs. chemotherapy in patients
- with NSCLCs that have a high-TMB (10mut/Mb)
First-line PD-1/PD-L1 +CTLA-4 in NSCLC
Summary
35. Outline
• Brief Introduction to Immunotherapy in NSCLC
• First-line Immunotherapy in NSCLC
• Immunotherapy for Stage III NSCLC
• Future Directions:
- Neoadjuvant Studies
- Translational Studies
36. Durvalumab in Stage III NSCLC
Phase III PACIFIC Trial: Study Design
Antonia et al, NEJM 2017
37. Antonia et al, NEJM 2017
Durvalumab in Stage III NSCLC
Phase III PACIFIC Trial: PFS
38. Durvalumab in Stage III NSCLC
Phase III PACIFIC Trial: Safety
Antonia et al, NEJM 2017
39. • Durvalumab Maintenance Therapy
- Is a new standard of care after definitive chemoradiation for
stage III NSCLC
Stage III NSCLC
Summary
40. Neoadjuvant Clinical Trials
General Principles
• Clinical and translational advantage over adjuvant approach
• No advances in resectable lung cancers since 2004
• Anti-PD-1/PD-L1 induces deep and durable responses in a
subset of patients with advanced NSCLC
• Neoadjuvant Anti-PD-1 may:
• Induce immunity against micrometastases
• Allow for a pre and post PD-1 pathologic assessment
• Provide tissue for correlative analyses
Forde et al, NEJM 2018
41. Slide 3
Neoadjuvant Nivolumab in Resectable NSCLC
Study Design
Primary Endpoint: Safety (drug-related adverse events 90-days post PD-1/30 days post surgery
Feasibility (resection without delay >37 days from pre-planned surgery)
Sample Size: 6 patient safety run-in, 20 patients in total
Exploratory Endpoints: Pathologic response, RFS, OS, immunologic correlates
Forde et al, NEJM 2018
42. Neoadjuvant Nivolumab in Resectable NSCLC
Patient Characteristics
Safety
-1 death in the postoperative period, unrelated
to study drug
Feasibility
- All patients underwent surgery without delay
- No complications in the postoperative period
22 enrolled
21 treated with
pre-op intent
1 withdrew
(SCLC histology)
20 resected
1 unresectable
Forde et al, NEJM 2018
43. Neoadjuvant Nivolumab in Resectable NSCLC
Response to Therapy
Radiologic Response after 2 doses
neoadjuvant nivolumab
Pleomorphic
NSCLC
Pathologic CR
Squamous
NSCLC
Pathologic CR
Pathologic Response after 2 doses
neoadjuvant nivolumab
Major Pathologic Response: <10% viable tumor
MPR in 43% (9/21) patients
Pre-nivo PD-L1 (>1%) did not correlate with MPR
Forde et al, NEJM 2018
44. Slide 14
Neoadjuvant Nivolumab in Resectable NSCLC
T-cell specific for dominant MANA
expand in peripheral blood
With thanks K. Smith and D. Pardoll
45. Slide 15
With thanks K. Smith and D. Pardoll
Neoadjuvant Nivolumab in Resectable NSCLC
T-cell specific for dominant MANA
expand in peripheral blood
46. Slide 16
With thanks K. Smith and D. Pardoll
Neoadjuvant Nivolumab in Resectable NSCLC
T-cell specific for dominant MANA
expand in peripheral blood
47. Neoadjuvant Nivolumab in Resectable NSCLC
TMB and Neoantigen density associates with MPR
With thanks K. Smith and D. Pardoll
48. Neoadjuvant Nivolumab in Resectable NSCLC
Conclusions
• Nivolumab prior to lung cancer resection did not delay time to surgery
• There were no unacceptable safety signals with neoadjuvant nivolumab
• 43% of tumors had MPRs
• Correlative analyses in a subset of tumors demonstrated
• Associations between mutation, MANA burden and pathologic response
• Identified MANA-specific TCRs in the blood and tumor
• Observed temporal increases in MANA-specific TCRs in the peripheral
blood post nivolumab: a potential biomarker of response
Forde et al, NEJM 2018
49. • Neoadjuvant chemoradiation for Resectable Stage IIIA NSCLC
- Used for a subset of patients with stage IIIA disease, deemed resectable
- Increased rate of pathologic CR, reduced rate of locoregional recurrence1,2
• Immunologic Synergy between PD-1/PD-L1 Blockade and RT
- Anti-CTLA-4 diversifies the TCR, which is associated with improved OS3
- Anti-PD-1+RT demonstrates TCR diversification in tumor and systemically4
Neoadjuvant Studies in Resectable NSCLC
Future Directions
1. Ettinger et al, JNCCN 2017
2. Albain et al, Lancet 2009
3. Demaria et al, Clin Cancer Res 2005
4. Deng et al, J Clin Investig 2014
4. Twyman-Saint Victor et al, Nature 2015
50. Neoadjuvant Immunoradiation
in Stage IIIA Resectable NSCLC
Cohort 2 opens when safety and feasibility of
cohort 1 established
IRB: J1772,
PI: Naidoo
51. JHH Thoracic Oncology Program
Julie R. Brahmer, MD
Patrick M. Forde, MD
David S. Ettinger, MD
Ronan J. Kelly, MD MBA
Christine L. Hann, MD PhD
Benjamin Levy, MD
Josephine Feliciano, MD
Kristen M. Marrone, MD
Jessica Wakefield, MS
Rachel Levy, MS
Joanne Riemer, RN
Cancer Biology Program
Valsamo Anagnostou, MD PhD
Victor E. Velculescu, MD PhD
Department of Radiology
Tony Lin, MD
Funding Sources
Stand Up To Cancer (SU2C)
Lungevity
IASLC
NIH KL2 program
JHH Bloomberg Kimmel Institute
for Cancer Immunotherapy (BKI)
Drew M. Pardoll, MD PhD
Elizabeth M. Jaffee, MD PhD
Kellie Smith, PhD
Suzanne Topalian MD
Department of Pathology
Tricia Cottrell, MD PhD
Janis Taube, MD
Peter Illei, MD
Colleagues and Collaborators
Editor's Notes
- Firstly, the interaction between a cancer and the immune system is a complex one, but can be simplified into a stepwise process called the cancer-immunity cycle.
- This starts with
Tumor ag release through cell death or destruction
Uptake and presentation of ag’s by APCs and the generation of an adaptive immune response and ag-specific t-cells in the LN
(priming and activation phase)
trafficking of ag-specific T-cells to the tme
and when they enter TME, recognize cancer cells as foreign via TCR specificity. When the TCR binds to the tumor cells that express known Ag, this induces a specific anti-tumor immune response.
This last step is tightly regulated by other inhibitory and activating factors on immune cells and tumor cells, that can potentially be therapeutically targeted.
and we can see that theorectically a number of anticancer agents can influence and potentially strengthen different stages of this cancer immunity cycle, in order to generate an immune mediated anti tumor response.
At the final stage in this cycle, therapeutic targeting of an inhibitory molecule that has garnered much interest is pd1 and it’s corresponding ligands PDL1 and PD-L2
- So what makes immunotherapy different to any other kind of therapy we hear about or is under investigation as an anti-cancer agent?
There are several unique features of the adaptive immune system in particular that makes it an appealing anti-cancer strategy,
Firstly, every T-cell is specific to a particular antigen
The immune system is able to evolve and generate new T cell clones in response to tumor evolution and antigen expression
Probably the most attractive factor, T-cells have memory- such that is a new antigen or antigen that elicits cross-reactivity with a particular T-cell clone exists, the adaptive immune system will maintain immunity against it, just as vaccines work for infective illnesses
And of course we all have an immune system, therefore these agents may not be specific tumor type, and may be able to work across tumor types, if an immune response against the tumor can be elicited
And indeed we have seen in clinical trials that clinical trials have provided proof of principle for many of these concepts, with anti-PD-1 therapy
- As this audience is aware, Dr. Brahmer led the first ph III clin trial of nivolumab compared to standard single agent docetaxel chemotherapy in the second-line squamous NSCLC in CM 017
In this study, patients had a 3m OS benefit, a 1m PFS benefit, and 21% of patients receiving nivo were progression free at 1 year compared to 6% of patients treated with docetaxel.
In a subset analysis based on PD-L1 status, it was found that all patients benefitted from nivo, regardless of PD-L1 expression
Hossein Borghaei lead a similar study in in the non-sq population in CM 057 study
The results of this study were very similar, in that we saw a 3m benefit in OS and 19% of patients were progression free at 1 yr compared to 8% with docetaxel
Pembro was studied in the large KEYNOTE 010 study in over 1000 pre-treated NSCLC patients
These studies included prospective validation of the PD-L1 IHC test using the 22c3 ab
In this study PD-L1 positive tumors again demonstrated an OS benefit in favor of pembro compared with chemo in second-line
In all patients, the HR for OS was 0.71 for pembro 2, 0.61 for pembro 10 compared with docetaxel
Additionally, about 1 third of patients were strongly expressors of pd-l defined as &gt;50% expression on tumor cells
In these patients the median OS was 14.9m for pembro 2, 17.3m for pembro 10 and 8.2m for docetaxel
This study demonstrated a benefit in OS in the first 850 pts, with an approx. 4m benefit in median OS in the non-sq population
Similarly, pts with sq NSCLC had a benefit in median OS of approx 1m, and HRs for benefit in OS was similar in both populations, and was ss
This study met it’s primary endpt, such that pembro demonstrated superior PFS (HR 0.50) over plt-doublet chemotherapy
Pembrolizumab also demonstrated superior OS vs. chemotherapy (HR 0.60)
In support of these findings the trial was stopped early for superior efficacy demonstrated in favor of pembro
• Pembrolizumab also provided a substantial improvement in ORR over chemotherapy: with 45% of patients responding compared with 28% with chemo
– This included 6 pts who were complete responders with pembrolizumab
- Data regarding DOR is still early, reflected by the median DOR NR in the pembro group, compared to 6.3m with docetaxel
- It is with this data in mind as well as a manageable safety profile, pembrolizumab has become a new standard of care for advanced NSCLC that expresses &gt;50% levels of PD-L1 in first-line
This is a pilot IIT run by my colleague Patrick Forde and myself in collaboration with MSKCC, where 20 pts with newly diagnosed resecteble stage IA-IIIA non-bulky NSCLC were given 2 doses of anti-PD-1 prior to surgery.
The co-primary endpoints of the study were
The study incorporated a 6 patient safety run in
And a host of correlative analyses were planned including pre and post Tx large vol blood draws, 6-8 core tissue biopsy samples and assessment of resection specimen and DLNs
This is a consort diagram of the study which demonstrates that 22 pts enrolled, one withdrew due to SCLC histology, 21 underwent preop tx, 20 proceeded to resection and 1 was unresectable
The patient characteristics are summarized in this table and are in keeping with a typical NSCLC population, median age 67, equal no. females and males, mostly ADC histology, and current/former smokers
Selected patients demonstrates deep responses to therapy with radiologic responses and pathologic complete responses, with evidence of residual viable tumor in the resection specimen after just 2 doses of nivolumab
The overall response results for the cohort enrolled are depicted in this waterfall plot, using the pathologic endpoint of MPR: defined as having &lt;10% viable tumor cells in the resection specimen, This has been used as a surrogate endpoint to clinical outcomes for neoadjuvant chemotherapy in nSCLC, and was used as an exploratory response endpoint here
In this study, 43%of patients demonstrated an MPR with nivo.
We also looked at whether pD-L1 status using the BMS antibody associated with response to rherapy, an using a cut-off of &gt;1% membranous staning as +. No association was seen between + PD-L1 status and MPR
Looking in closer detail at the MANAfest assays performed in one particular patient enrolled in the neoadjuvant study, 3 T-cell clones were identified that corresponded to a particular MANA within the tumor tissue prior to therapy
Interestingly, the MANA-specific t-cell clones of interest were only found to comprise 1.6% of the initial tumor sample from primary tumor and LN biopsy prior to neoadjuvant therapy
- In this study, we observed dynamic changes in the 3 T-cell clones linked to this MANA in response to anti-PD-1, that expanded in the peripheral blood prior to surgery
Coming back to our neoadjuvant trial of nivo in NSCLC, we saw a positive association between the number of nonsynonymous tumor mutations and MANA density and the development of MPR, regardless of tumor histology
And regardless of pd-l1 status
Therefore in summary
- From here, we hope to expand on these observations and have expanded the previously mentioned study to evaluate the combination of ipi/nivo in stage IB-IIIA non-bulky patients
- In addition, a subset of patients with stage IIIA NSCLC have bulky mediastinal LN and may be treated with trimodality therapy involving chemo,RT and surgery.
We also know from preclinical studies that radiation therapy may support anti-tumor immune responses through diversification of the TCR, and in some instances this has correlated with improved clinical outlcomes
Therefore we will be opening a follow-up study of the combination of an anti-PD-1 antibody durva together with RT and a second cohort exploring the combination of durva/chemo and RT prior to surgery in patients with stage IIIA disease that is amenable to surgical resection.
We will aim to perform similar integrated genomic and immunologic analyses in these patients including MANAfest
- Again thank you to the clinical and translational team at hopkins, and to the conference
Organizers, and I would be glad to take questions or hear comments.