Dr. Jack West's presentation on highlights in advanced non-small cell lung cancer from ASCO 2014, focusing on new agents ramucirumab and necitumumab for broad NSCLC populations, crizotinib and ceritinib for ALK-positive NSCLC, EGFR inhibitor-options of afatinib and bevacizumab added to erlotinib for first line treatment of EGFR mutation-positive NSCLC, and AZD9291 or CO1686 for EGFR mutation-positive patients with acquired resistance.

1. Best of ASCO 2014:
Highlights in Metastatic
Non-Small Cell Lung Cancer
Howard (Jack) West, MD
@JackWestMD
Swedish Cancer Institute
Seattle, WA
Best of ASCO 2014
Seattle, WA

2. Learning Objectives
• Determine whether evidence on necitumumab and
ramucirumab for broad NSCLC populations are sufficient to
modify current treatment standards
• Evaluate treatment options of crizotinib and ceritinib for
ALK-positive advanced NSCLC
• Compare utility of various EGFR TKI-based options as first
line treatment of advanced EGFR mutation-positive NSCLC
• Recognize efficacy of both AZD9291 and CO1686 in
treating acquired resistance to EGFR TKIs (*T790M-
positive)

3. Novel Agents in
Broad (Molecularly Unselected)
Advanced NSCLC Populations

4. SQUIRE: Chemotherapy +/- Necitumumab
for First Line Adv Squamous NSCLC
Adv squamous
NSCLC
Treatment-naïve
N = 1093
R
A
N
D
O
M
I
Z
E
Cisplatin/Gemcitabine
+ Necitumumab
up to 6 cycles
Primary endpoint: OS
• Necitumumab (Neci) (IMC-11F8) is a human IgG1 anti-EGFR
monoclonal antibody
Cisplatin/Gemcitabine
up to 6 cycles
Maintenance
neci until
progression
Cisplatin 75 mg/m2 IV day 1 q21 days
Gemcitabine 1250 mg/m2 IV days 1, 8 q21 days
Necitumumab 800 mg/kg IV days 1, 8 q21 days
Thatcher, A#8008

5. SQUIRE: Efficacy of Necitumumab
Overall Survival (ITT)Progression-Free Survival (ITT)
Thatcher, A#8008
Chemo/Neci
(N = 545)
Chemo alone
(N = 548)
P
ORR (CR + PR) 31.2% 28.8% 0.400
DCR (CR + PR +SD) 81.8% 77.0% 0.043*
*Cochran-Mantel-Haenszel test (stratified)

6. SQUIRE: Toxicity of Necitumumab
Thatcher, A#8008

7. SQUIRE: FLEX Redux? (Re-FLEX?)
SQUIRE FLEX
• Extremely similar agent; extremely similar results
– cetuximab has had negligible impact on NSCLC management
• Highlights distinction between statistical & clinical significance
Pirker, Lancet 2009Thatcher, A#8008

8. REVEL: Docetaxel +/- Ramucirumab as
Second Line Therapy for Adv NSCLC
Adv NSCLC
(any histology)
Prior platinum-
based chemo
Prior bev allowed
N = 1253
R
A
N
D
O
M
I
Z
E
Docetaxel 75 mg/m2 +
Ramucirumab 10 mg/kg
IV Q21 days
Docetaxel 75 mg/m2 +
Placebo
IV Q21 days
Primary endpoint: OS
Treat until PD
or prohibitive
toxicity
• Ramucirumab (RAM) is a human IgG1 monoclonal antibody,
specifically binding to the extracellular domain of VEGFR-2
• Approved in previously treated gastric cancer
Perol, A#LBA-8006

9. REVEL: Efficacy of Ramucirumab
Overall Survival (ITT)Progression-Free Survival (ITT)
RAM + DOC
(N = 628)
PL + DOC
(N = 625)
P
ORR (CR + PR) 22.9% 13.6% <0.001
DCR (CR + PR +SD) 64.0% 52.6% <0.001

10. REVEL: Toxicity (Adverse Events in
>20% of Patients or >5% Higher w/RAM

11. Half Empty or Half Full?
• 1.4 mo increase in OS isn’t much
• Cost: about $7K/treatment
• But options improving OS >2nd line
are very limited, especially for
squamous NSCLC
Optimistic oncologist perspective
• Whether adding new agent with some increased toxicity
and additional significant cost is “worth it” for 1.4 mo
improvement in median OS is your judgment
• Not a clear change in standard of care

12. (Not Very) Old and New
Agents for ALK-Positive
Advanced NSCLC

13. PROFILE 1014: First-Line Crizotinib vs.
Chemo in ALK-Positive Adv NSCLC
• Crizotinib has significant activity, RR 60-65%, in previously
treated ALK+ NSCLC, & PFS 7.7 mo vs. 3.0 months
compared with 2nd line chemo (Shaw, ESMO 2012, A#2862)
• Though largely presumed to be superior to first line chemo
in ALK-positive NSCLC, this hasn’t been prospectively
demonstrated
Advanced
NSCLC ALK+
No Prior Rx
N= 343
R
A
N
D
Crizotinib
Cisplatin/Pemetrexed or
Carboplatin/Pemetrexed
Primary endpoint: PFS
1: 1
Mok, A#8002

14. First-Line Crizotinib vs. Chemo:
Progression-Free Survival
Mok, A#8002

15. Responses, Crizotinib vs. Chemo
74%
45%
Mok, A#8002
Criz Chemo
ORR 74% 45%
Resp Dur (wks) 49 23

16. Common AEs of Any Cause in ≥25% of Patients
With ≥5% Difference Between Groupsa
Crizotinib (n=171), n (%) Chemotherapy (n=169), n (%)b
Any grade Grade 3/4 Any grade Grade 3/4
Higher frequency (≥5% absolute difference) in crizotinib arm
Vision disorderc 122 (71) 1 (1) 24 (14) 0
Diarrhea 105 (61) 4 (2) 29 (17) 1 (1)
Edemac 83 (49) 1 (1) 22 (13) 1 (1)
Vomiting 78 (46) 3 (2) 68 (40) 6 (4)
Constipation 74 (43) 3 (2) 53 (31) 0
Elevated transaminasesc 61 (36) 24 (14) 22 (13) 4 (2)
Upper respiratory infectionc 55 (32) 0 21 (12) 1 (1)
Abdominal painc 45 (26) 0 20 (12) 0
Dysgeusia 45 (26) 0 11 (7) 0
Higher frequency (≥5% absolute difference) in chemotherapy arm
Nausea 95 (56) 2 (1) 103 (61) 3 (2)
Decreased appetite 51 (30) 4 (2) 59 (35) 1 (1)
Fatigue 49 (29) 5 (3) 65 (38) 4 (2)
Neutropeniac 36 (21) 19 (11) 51 (30) 26 (15)
Asthenia 22 (13) 0 42 (25) 2 (1)
Anemiac 15 (9) 0 54 (32) 15 (9)
a
Not adjusted for differential treatment duration; bbefore crossover to crizotinib; cclustered term
● Permanent treatment discontinuations due to treatment-related AEs: 5% in crizotinib arm;
8% in chemotherapy armb
● No grade 5 AEs reported to be related to treatment; 1 patient in chemotherapy arm had grade 5
pneumonitis after crossover to crizotinib, considered to be treatment-related

17. Ceritinib: New Treatment
Option for ALK-Positive NSCLC
Kim, A#8003
FDA Approved
April, 2014
Cost:
$13,500/mo

18. #
##
Dose Expansion Cohort, 750 mg/day:
Best Percentage Change from Baseline
NSCLC with prior ALKi
NSCLC ALKi naïve
Changefrombaselineinsumoflongestdiameters(%)
*Patients with measurable disease at baseline and at least 1 post baseline assessment
without unknown response for target lesion or overall response
N=228*
#
#
#
#
#
#
#
#
#
#
#
## # # # #
#
# # #
# # #
# # # #
#
# # # # # # # #
# # #
# # #
#
#
# #
#
#
#
#
## ####
#
#
#
##
# #
# #
#
# #
#
###
##
#
#
#
#
#
#####
##
##
# #
##
# ###
#######
#
#
# ###
# #
#
###
#
#
#PFS event
Kim, A#8003

19. Activity of Ceritinib in ALK+ NSCLC
Endpoint Criz-Naïve
(N = 83)
Criz-Refractory
(N = 163)
Response Rate 66% 55%
12-mo Prog-Free Survival 61% 28%
Median Time from Dx to
Ceritinib
8.1 mo 21.2 mo
Disease Control, Brain Mets 70% 75%
• Modestly higher RR, longer responses in crizotinib-
naïve patients
Kim, A#8003

20. Toxicity Challenges with Ceritinib
• Greater than with crizotinib
• Dose reduction – 59% (!)
– Increased ALT/AST, nausea, diarrhea, vomiting
• Discontinuation due to adverse effects – 10%
– Pneumonia, ILD/pneumonitis, decreased appetite
• Oncologists need to know to dose-reduce early
– 750 mg daily may be more than needed
Kim, A#8003

21. Timing of Ceritinib?
Presented by: H. Jack West
• Approval is for crizotinib-refractory and crizotinib-intolerant
patients with ALK rearrangement
• Should it be used earlier?
ALK+
No Prior Rx
N= 348
R
A
N
D
Ceritinib
Cisplatin/Pemetrexed or
Carboplatin/PemetrexedPrimary endpoint: PFS
Trial in
development
ALK+
No Prior Rx
R
A
N
D Ceritinib
Crizotinib
Primary endpoint: OS
Trial
we need
Ceritinib
Chemo or MD choice

22. Assessing Options for
First Line Treatment of
EGFR Mutation-Positive
Advanced NSCLC

23. LUX Lung-3, LUX Lung-6 Trials
EGFR Mut’n Pos
Advanced NSCLC
No Prior Rx
N= 345
Global
R
A
N
D
Afatinib 40 mg PO daily
until progression
Cisplatin/Pemetrexed Q21d
up to 6 cycles
Primary endpoint: PFS
Sequist, JCO 2013LUX Lung-3
R
A
N
D
Afatinib 40 mg PO daily
until progression
Cisplatin d1, Gemcitabine d1,8 q21d
up to 6 cycles
Wu, Lancet 2014
EGFR Mut’n Pos
Advanced NSCLC
No Prior Rx
N= 364
Asia
Primary endpoint: PFS
LUX Lung-6
2:1
2:1

24. OS Analysis: LUX-Lung 3, LUX-Lung 6
(Del 19 and L858R only)
Yang, A#8004
1.0
0.8
0.6
0.4
0.2
0
EstimatedOSprobability
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51
Time (months)
203 197 188 181 171 162 143 133 121 108 101 90 58 49 32 9 1 0
104 98 92 86 81 71 63 55 52 47 40 35 26 20 10 5 1 0
Afatinib
Pem/Cis
No of patients
LUX-Lung 3
Afatinib
n=203
Pem/Cis
n=104
Median,
months 31.57 28.16
HR (95%CI),
p-value
0.78 (0.58–1.06),
p=0.1090
LUX-Lung 6
Afatinib
n=216
Gem/Cis
n=108
Median,
months 23.6 23.5
HR (95%CI),
p-value
0.83 (0.62–1.09),
p=0.1756
1.0
0.8
0.6
0.4
0.2
0
EstimatedOSprobability
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45
Time (months)
216 214 202 190 172 158 141 118 104 93 80 51 19 9 1 0
108 101 93 87 81 70 61 55 49 36 30 17 8 3 0 0
Afatinib
Gem/Cis
No of patients

25. Combined OS Analysis: LUX-Lung 3,
LUX-Lung 6 (Del 19 and L858R only)
Yang, A#8004

26. Combined OS Analysis: LUX-Lung 3,
LUX-Lung 6 by Mutation Subtype
Yang, A#8004

27. Differences in Efficacy of Gefitinib/ Erlotinib:
Exon 19 Del vs. L858R
OSPFS
Jackman,
Clin Cancer Res,
2006
Riely,
Clin Cancer Res,
2006
OS

28. Treatment after Progression on First Line
Therapy (Del 19 and L858R only)
LUX-Lung 3 LUX-Lung 6
Afatinib
(n=203)
Pem/Cis
(n=104)
Afatinib
(n=216)
Gem/Cis
(n=108)
Discontinued treatment, n (%) 184 (100) 104 (100) 194 (100) 108 (100)
Subsequent systemic therapy, n (%)† 144 (78) 88 (85) 123 (63) 70 (65)
Chemotherapy, n (%) 131 (71) 49 (47) 114 (59) 29 (27)
EGFR TKI therapy, n (%)
Erlotinib
Gefitinib
Afatinib
AZD9291
Dacomitinib
Icotinib
EGFR TKI combinations
81 (44)
61 (33)
28 (15)
2 (1)
2 (1)
–
–
5 (3)
78 (75)
46 (42)
44 (42)
7 (7)
1 (1)
1 (1)
–
9 (9)
50 (26)
21 (11)
19 (10)
–
–
–
11 (6)
5 (3)
61 (56)
22 (20)
39 (36)
–
–
–
3 (3)
3 (3)
Other systemic therapy±, n (%) 5 (3) 2 (2) 3 (2) 4 (4)
Radiotherapy, n (%) 32 (17) 21 (20) 4 (2) 0 (0)
†Collection of data on subsequent therapies still ongoing.
± include investigational agents, monoclonal antibodies, non-EGFR targeting protein kinase inhibitors etc
Yang, A#8004

29. Treatment after Progression on First Line by
Country’s Reimbursement*
Countries with universal
reimbursement policies**
Countries without
universal reimbursement
policies***
Afatinib
(n=144)
Chemo
(n=75)
Afatinib
(n=275)
Chemo
(n=137)
Discontinued treatment, n (%) 127 (100) 75 (100) 251 (100) 137 (100)
Subsequent systemic therapy, n (%) 112 (88) 69 (92) 158 (63) 89 (65)
Chemotherapy, n (%) 103 (81) 35 (47) 142 (57) 43 (31)
EGFR TKI, n (%) 76 (60) 68 (91) 55 (22) 71 (52)
Other, n (%) 5 (4) 2 (3) 3 (1) 4 (3)
Radiotherapy, n (%) 27 (22) 18 (24) 9 (4) 3 (2)
*Determined by presence or absence of a national reimbursement policy in effect throughout the period of trial conduct:
**Main countries contributing : Japan, Taiwan, Korea, Germany, France, Australia, UK, Belgium
***Main countries contributing : China, Thailand, Russia, the Philippines, Malaysia
Yang, A#8004

30. Impact of subsequent EGFR TKIs on OS:
exploratory analyses by category of EGFR TKI
reimbursement policy in country of residence*
Population % received
TKI after 1st
line chemo
HR (95% CI)
Common
mutations
HR (95% CI)
Del19
HR (95% CI)
L858R
Combined LL3/6 population
(n=631)
66% 0.81
(0.66–0.99)
0.59
(0.45–0.77)
1.25
(0.92–1.71)
Countries with universal
reimbursement policies**
(n=219)
91% 0.71
(0.49–1.02)
0.50
(0.31–0.81)
1.14
(0.64–2.03)
Countries without universal
reimbursement policies***
(n=412)
52% 0.85
(0.66–1.08)
0.59
(0.42–0.82)
1.32
(0.91–1.92)
Japan
(n=77)
100% 0.57
(0.29–1.12)
0.34
(0.13–0.87)
1.13
(0.40–3.21)
*Determined by presence or absence of a national reimbursement policy in effect throughout the period of trial conduct:
**Main countries contributing : Japan, Taiwan, Korea, Germany, France, Australia, UK, Belgium, Ireland
***Main countries contributing : China, Thailand, Russia, the Philippines, Malaysia
Yang, A#8004

31. Is OS with Afatinib in LUX-Lung 3/6
Superior to That of Other EGFR TKIs?
Med
OS(mo)
WJTOG2 EURTAC3 OPTIMAL4 LUX-
Lung 35
NEJGSG1
Gefitinib Erlotinib Afatinib
1) Maemondo et al. N Engl J Med. 2010;362:2380; 2) Mitsudomi et al. Lancet Oncol. 2010;11:121; 3) Rosell et al.
Lancet Oncol. 2012;13:239; 4) Zhou et al. Lancet Oncol. 2011;12:735; 5) Yang et al. Proc ASCO 2014: A#8004
N=230 N=177
Terminated
early, after
Interim analysis?
N=174 N=165 N=345 N=364
LUX-
Lung 65
EGFR TKI
Chemo

32. Multiple Shots on Goal
• No survival difference for
overall ITT population
• No survival difference for
LUX-Lung trials individually
• Statistical significance is
function of magnitude of
difference & population size
• Pooling two trials and eliminating rare mutations 
statistical significance
• Still, OS benefit for Del19 pts is robust, impressive

33. Relevant Comparison for Afatinib
in 2014 is to Other EGFR TKIs
• Is timing of EGFR TKI critical re: crossover?
– L858R population showed PFS benefit but reversal w/OS
– Sequence of therapy may be relevant
• Would other EGFR TKIs show OS benefit if > 700
pts enrolled & results divided by mut’n subtype?
EGFR Mut+
N = 316
(Asia)
R
A
N
D Afatinib daily
Gefitinib daily
Completed July, 2013Primary endpoint: OS
LUX-Lung 7
• Toxicity assessment will also be critical

34. Or is optimal therapy now an
EGFR TKI/bevacizumab
combination?

35. Subsets Demonstrate Benefit with Bev
Added to 2nd Line Erlotinib (BeTa)
Herbst, Lancet 2011

36. Erlotinib +/- Bevacizumab as Maintenance
Therapy (ATLAS): Clinical Subgroups
Johnson, JCO 2013

37. EGFR Regulates VEGF in
EGFR Mutant Cell Lines
Courtesy of Heymach et al.; manuscript in preparation
0
500
1000
1500
2000
2500
VEGF(pg/300000cell)
EGF(60ng/ml) - - + + - - + + - - + + - - + +
Erlotinib(1mM) - + - + - + - + - + - + - + - +
A549 (wt)
HCC827 (mut)
H3255 (mut)
H1975 (mut)
wild-type mutant Mutant + T790M
• EGFR mutant NSCLC cell lines express higher levels of VEGF
• VEGF expression is reduced with EGFR inhibitors

38. Erlotinib/Bevacizumab vs. Erlotinib for
EGFR Mutation-Positive Adv NSCLC
Adv NSCLC
EGFR Mut’n (exon 19/21)
Treatment-naïve
N = 154
R
A
N
D
Erlotinib 150 mg/day
+ bevacizumab 15 mg/kg IV Q21 days
until progression or prohibitive toxicity
Primary endpoint: PFS
Erlotinib 150 mg/days
until progression or prohibitive toxicity
EB E P
ORR (CR/PR) 69% 64% ns
DCR (CR/PR/SD) 99% 88% 0.018
Kato, A#8005

39. JO25587 PFS in Context of Other Trials
in EGFR Mutation-Positive NSCLC
1. Rosell et al. Lancet Oncol. 2012;13:239; 11. Zhou et al. Lancet Oncol. 2011;12:735; 3. Sequist et al. J Clin Oncol. 2013;31:3327; 4.
Wu et al. Lancet Oncol. 2014;15:213; 5. Kato, Proc ASCO 2014, A#8005
Med
PFS (mo)
EURTAC1 OPTIMAL2 JO25587-
Erlotinib5
JO25587-
Erloti/Bev5
LUX-
Lung-33
LUX-
Lung-64

40. Erlotinib +/- Bevacizumab:
PFS by EGFR mutation type
EB group E group
Median (months) 18.0 10.3
HR 0.41
(95% CI: 0.24–0.72)
EB
E
E
EB
Number at risk
0
1.0
0
40
40 E
EB
Number at risk
EB
E
1.0
0
0
35
37
Time (months)
4 8 122 6 10 14 18 22 2616 20 24 28
38 33 2739 36 29 24 12 5 219 8 2 0
29 22 1235 26 16 9 5 1 09 3 0 0
Time (months)
4 8 122 6 10 14 18 22 2616 20 24 28
31 27 2233 28 24 14 8 3 211 5 2 0
28 17 1231 18 13 12 7 4 19 7 2 0
0.2
0.4
0.6
0.8
PFSprobability
PFSprobability
0.2
0.4
0.6
0.8
Exon 19 deletion Exon 21 L858R
EB group E group
Median (months) 13.9 7.1
HR 0.67
(95% CI: 0.38–1.18)
Kato, A#8005

41. Toxicity Issues with Erlotinib/Bevacizumab
on JO25587
• No unforeseen toxicities or Rx-related deaths
• Grade >3 toxicity 91% vs. 53% (esp. HTN, proteinuria)
• 41% discontinued bevacizumab for adverse effects
– Primarily proteinuria (15%) or hemorrhagic (12%)
• Bevacizumab discontinuation rate 10-15% in
BeTa, ATLAS trials
• Difference?
– Greater toxicity in Japanese population?
– Greater toxicity in EGFR mutation-positive?
– Longer duration of therapy  higher risk of ADRs

42. Cost Considerations with
Erlotinib/Bev Combination
Cost/
Month
($USD)
$6,300
$16,700
Erlotinib Erlotinib/
Bevacizumab
Addition of bevacizumab increases cost of first line
treatment by ~$120,000 for 16 treatments
(acquisition cost alone)

43. Is Erlotinib/Bevacizumab the
New Standard of Care for EGFR Mut+?
• Kato: median PFS 16.0 vs. 9.7 mo, HR 0.54
(ECOG 4599: median PFS 6.2 vs. 4.5 mo, HR 0.66)
• My preferred regimen moving forward
• Threshold for clinical confidence ≠ threshold for coverage
• Confirmatory trial needed? Outside of Japan? OS diff?
• Will it be covered?
EGFR Mut+
N = 118
N Amer
R
A
N
D
Erlotinib daily
Erlotinib daily &
Bevacizumab q3wk
Ongoing
(slowly)
Primary endpoint: PFS
ACCRU Trial
PI – T. Stinchcombe

44. Breaking the Impasse for
EGFR Mutation-Positive Patients
with Acquired Resistance

45. Acquired Resistance:
AZD9291 and CO-1686
• The vast majority of patients with an activating EGFR
mutation who respond well to initial EGFR TKI therapy
demonstrate progression several months to years later
• T790M mutation detected in approximately 60% of
patients with acquired resistance
• AZD9291 and CO1686 are mutant selective,
irreversible inhibitors of EGFR with significant anti-
tumor activity in preclinical tumor models with both
EGFR TKI-sensitizing and T790M resistance mutations
with greater wild-type sparing than first generation
EGFR TKIs

46. Best percentage change from baseline in target lesion:
all evaluable patients, escalation and expansion (N=205)
Jänne, A#8009
AZD9291: Response rate*
in overall population( T790M+ and T790M-)
• First patient dosed Mar 6, 2013
• Longest response >9 months ongoing at time of data cutoff
• ORR* = 53% (109/205; 95% CI 46%, 60%); no difference in ORR by race
• Overall disease control rate (CR+PR+SD) = 83% (171/205; 95% CI 78%, 88%)
20 mg 40 mg 80 mg 160 mg 240 mg
N (205) 20 57 61 55 12
ORR 55% 44% 54% 58% 67%
*Includes confirmed responses and responses awaiting confirmation; #represents imputed values. Population: all dosed patients with a
baseline RECIST assessment and an evaluable response (CR, PR, SD, or PD), N=205 (from 232 dosed patients, 27 patients with a current
non-evaluable response are not included). CI, confidence interval; CR, confirmed complete response; ORR, overall response rate;
PD, progressive disease; PR, confirmed partial response; RECIST, Response Evaluation Criteria In Solid Tumors; SD, stable disease
40
20
0
-20
-40
-60
-80
-100
Complete response
#######
Partial response*
Non-response

47. Jänne, A#8009
AZD9291: Response rate* in T790M+
(central test)
• ORR* = 64% (69/107; 95% Cl 55%, 73%) in patients with EGFR T790M+ NSCLC
• Overall disease control rate (CR+PR+SD) = 94% (101/107; 95% CI 88%, 98%)
20 mg 40 mg 80 mg 160 mg 240 mg
N (107) 10 29 34 28 6
ORR 50% 62% 68% 64% 83%
Best percentage change from baseline in target lesion: all evaluable T790M+ patients, Part B
*Includes confirmed responses and responses awaiting confirmation; #represents imputed values
Population: all dosed centrally confirmed T790M+ patients with a baseline RECIST assessment and an evaluable response (CR/PR, SD, or PD),
N=107 (from 120 T790M+ patients; 13 patients with a current non-evaluable response are not included). D, discontinued; QD, once daily
Best percentage change from baseline in target lesion:
T790M+ evaluable patients, expansion cohorts only (N=107)
40 mg QD
80 mg QD
160 mg QD
240 mg QD
20 mg QD
40
20
-20
-40
-60
-80
-100
0
# #
D D
D D
D
D DD D
D D
D D
D D
D
D D
D D
D D
D

48. • ORR* = 22% (11/50; 95% Cl 12%, 36%) in patients with EGFR T790M- NSCLC
• Overall disease control rate (CR+PR+SD) = 56% (28/50; 95% CI 41%, 70%)
20 mg 40 mg 80 mg 160 mg
N (50) 3 17 17 13
ORR 67% 12% 24% 23%
*Includes confirmed responses and responses awaiting confirmation; #represents imputed values
Population: all dosed centrally confirmed T790M- patients with a baseline RECIST assessment and an evaluable response (CR/PR, SD, or PD),
N=50 (from 56 T790M- patients; six patients with a current non-evaluable response are not included)
40 mg QD
80 mg QD
160 mg QD
20 mg QD
# # # #
D D D D
40
20
-20
-40
-60
-80
-100
0
D
D
D
D D
D
D
D D D
D D D D D D D D
D D D
D
D
D
AZD9291: Response rate* in T790M-
(central test)
Jänne, A#8009

49. T790M+ T790M-
68/105
43/69
25/36
11/50
3/28
8/22
Response rate* according to T790M (central test)
status: immediate prior EGFR-TKI,# yes vs no
*Includes confirmed responses and responses awaiting confirmation; #TKI therapy is defined as being immediately
prior if TKI was the last regimen taken prior to the study, with no subsequent therapy. Population: all dosed centrally confirmed
T790M+ and T790M- patients with a baseline RECIST assessment and an evaluable response, T790M+ N=105 (from 107 T790M+
patients with response data; two patients not included as subgroup missing), T790M- N=50
Jänne, A#8009

50. AZD9291: Progression-free survival by
T790M (central test) status
0 6 12 18 24 30 36 42
Study week
Probabilityofprogression-freesurvival
T790M+ (95% CI)
T790M- (95% CI)
Dots indicate censored observations, shaded area represents 95% CIs; progression events that do not occur within
14 weeks of the last evaluable assessment (or first dose) are censored. Population: all dosed centrally confirmed T790M+
and T790M- patients, N=170 (115 T790M+, 55 T790M-; six patients for whom start date is not yet known are not included)
0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Patients at risk
T790M+
T790M-
115
55
96
33
78
21
56
15
21
7
6
0
1
0
Jänne, A#8009

51. Jänne, A#8009
AZD9291: Toxicity Summary
• No clear maximum-tolerated dose up to 240 mg/d
– No clear correlation of toxicity with dose
– 80 mg/d selected as optimal for therapeutic index
• At 80 mg/d dose:
– Diarrhea in 20%
– Rash in 27%
– Interstitial lung disease in 3%
– Hyperglycemia 1%

52. CO-1686: Best Response in Phase I
and Early Phase II Cohort Patients
Sequist, A#8010

53. CO-1686: Progression-Free Survival
Sequist, A#8010

54. Sequist, A#8010
CO-1686: Toxicity Profile

55. Current/Future Development of Third
Generation EGFR TKIs
• AZD9291
– AURA: Ph 2 expansion in T790M+
– AURA 2: Confirmatory Ph 2 in T790M+
– AURA 3: Ph 3 2nd line vs. chemo in T790M+
• CO-1686
– TIGER X: expansion of 2nd line or later in T790M+
– TIGER 1: Ph 2/3 1st line vs. erlotinib
(no T790M+ requirement)
– TIGER 2: Ph 2 after 1 prior EGFR TKI in T790M+
– TIGER 3: Ph 2 vs. chemo in T790M+

56. Summary of Third Generation
EGFR TKIs in Acquired Resistance
• BOTH agents have impressive efficacy and are
granted breakthrough designation by FDA
– Foot race to clinic
• Toxicity differences:
– AZD9291: rash & diarrhea (<erlotinib), ILD in minority
– CO-1686: hyperglycemia, mild diarrhea, nausea, rare
QTc prolongation
• Though some potential toxicity concerns in both,
anticipated benefit >> risk
• For both agents, focus is on T790M+ patients

57. Closing Thoughts/
Summary

58. No absolute rule for the amount of
evidence required to change practice
• Big effects don’t require big trials
• Crizotinib was justifiably approved based
on phase I/II trial with >50% RR
– Ceritinib similarly approved despite relatively small
numbers, but large effect
• EGFR TKI therapies became clear first line
standard of care despite absence of OS benefit
– Different standard for adding bevacizumab?
• Conversely, is an intervention clinically significant
if the trial needs >1000 patients to demonstrate
statistical significance?

59. In 2014, Cost/Value of Therapy is a
Factor in Cancer Care
• Reality is that cost matters, especially as new
drugs have eclipsed the prior $10,000/mo barrier
• It is appropriate to expect a semblance of value
and not merely p < 0.05
• We need to discuss value openly and not just
have it bias our clinical judgment
• Cost is limiting our ability to deliver best treatment
Optimal Rx
($$$$) Cost/practical
limits
Drug delivery
to needy patients

60. Take Home Messages for Advanced
NSCLC Management from ASCO 2014
• Necitumumab: Re-FLEX?; not enough benefit vs. toxicity
• Ramicirumab for 2nd line: Maybe; is 1.4 mo med OS diff
worth cost?; improving OS is hard, esp in squam NSCLC
• Crizotinib >> chemo first line (RR & PFS) in ALK+
• Ceritinib highly active for criz-naïve or criz-resistant
ALK+(RR & PFS), & CNS activity, but toxicity challenging
• Afatinib OS benefit specific to Del 19; Unique to afatinib?
– Major differences between Del 19 and L858R populations
• Bevacizumab significantly improved PFS w/erlotinib for
EGFR mut’n pos-NSCLC
• AZD9291 and CO-1686 both very active in EGFR T790M+
acquired resistance after prior 1st gen EGFR TKI