Transcript ASCO_GI_2011_files/Berlin HER family gastroesoph ASCO

The HER Superfamily of
Receptors and Related
Pathways in
Gastroesophageal Cancers
AKA: ToGA: The starting point
Jordan Berlin, M.D.
Ingram Associate Professor of
Medicine
Vanderbilt-Ingram Cancer
Center
Disclosures
 For this talk
– Advisory board and research support:
Roche/Genentech
– Advisory Board and research support:
Amgen
– Research Support: Lilly/Imclone
 I have a lot of relationships, but
no others relevant
ToGA trial design
Phase III, randomized, open-label, international, multicenter study
3807 patients screened1
810 HER2-positive (22.1%)
HER2-positive
advanced GC
(n=584)
5-FU or capecitabinea
+ cisplatin
(n=290)
R
5-FU or capecitabinea
+ cisplatin
+ trastuzumab
(n=294)
 Stratification factors
−
−
−
−
−
advanced vs metastatic
GC vs GEJ
measurable vs non-measurable
ECOG PS 0-1 vs 2
capecitabine vs 5-FU
1Bang
aChosen
at investigator’s discretion
GEJ, gastroesophageal junction
et al; Abstract 4556, ASCO 2009
Rationale for trastuzumab in
HER2-positive GC
 There is no universal standard treatment, but
– fluoropyrimidine (capecitabine / 5-FU) / platinum
(cisplatin / oxaliplatin)-based chemotherapy considered
as reference regimen
• epirubicin or docetaxel sometimes added
• biologicals are under investigation
 Unmet need for new treatment options in advanced
gastric cancer
 Some gastric adenocarcinomas are HER2 positive
 Trastuzumab is effective against HER2-overexpressing
GC cell lines in vitro and in vivo
Fujimoto-Ouchi et al 2007; Gravalos & Jimeno 2008
Primary end point: OS
Event
Median
Events OS
HR
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
FC + T
FC
11.1
0
2
4
6
167
182
13.8
11.1
95% CI
p value
0.74 0.60, 0.91
13.8
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
Time (months)
No.
at risk
294 277 246 209 173 147 113 90
290 266 223 185 143 117 90 64
T, trastuzumab
71
47
56
32
43
24
30
16
21
14
13
7
12
6
6
5
4
0
1
0
0
0
0.0046
Secondary end point: PFS
Event
Median
Events PFS HR
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
FC + T
FC
5.5
0
2
4
226
235
6.7
5.5
95% CI
p value
0.71 0.59, 0.85
6.7
6
8 10 12 14 16 18 20 22 24 26 28 30 32 34
Time (months)
No.
at risk
294 258 201 141 95
290 238 182 99 62
60
33
41
17
28
7
21
5
13
3
9
3
8
2
6
2
6
1
6
1
4
0
2
0
0
0
0.0002
OS in IHC2+/FISH+ or IHC3+
(exploratory analysis)
Event
Median
Events OS
HR
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
FC + T
FC
11.8
0
2
4
6
120
136
16.0
11.8
95% CI
0.65 0.51, 0.83
16.0
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
Time (months)
No.
at risk
228 218 196 170 142 122 100 84
218 198 170 141 112 96 75 53
65
39
51 39
28 20
28
13
20 12
11 4
11
3
5
3
4
0
1
0
0
0
CALGB 80403 / ECOG E1206: Schema
ARM A: (ECF + cetuximab); 1 cycle = 21 days
Cetuximab 400  250mg/m2 IV, weekly
Epirubicin 50 mg/m2 IV, day 1
Cisplatin 60mg/m2 IV, day 1
Fluorouracil 200mg/m2/day, days 1-21
Stratification:
ECOG 0-1 vs 2
ADC vs. SCC
ARM B: (IC + cetuximab); 1 cycle = 21 days
Cetuximab 400  250mg/m2 IV, weekly
Cisplatin 30 mg/m2 IV, days 1 and 8
Irinotecan 65 mg/m2 IV, days 1 and 8
ARM C: (FOLFOX + cetuximab); 1 cycle = 14 days
Cetuximab 400  250mg/m2 IV, weekly
Oxaliplatin 85 mg/m2 IV, day 1
Leucovorin 400 mg/m2, day 1
Fluorouracil 400 mg/m2 IV bolus, day 1
Fluorouracil 2400 mg/m2 IV over 46hrs (days 1-2)
Enzinger PC and Burtness B, et al. ASCO 2010
CALGB 80403 / ECOG E1206: Background
•
Cetuximab: chimerized monoclonal antibody EGFR (oropharyngeal cancer , NSCLC, and
colorectal cancer)
•
EGFR expression – 3/4 of ADC and SCC 1-5
•
EGFR expression correlates with prognosis in
esophagogastric ADC and SCC 1-5
•
KRAS mutations occur in approx. 2% of
esophageal cancers6
1-5 Mukaida.
Cancer 1991; Itakura. Cancer
1994; Yacoub. Mod Pathol 1997; Torzewski.
Anticancer Res 1997; Koyama. J Cancer Res
Clin Oncol 1999; 6 Lea. Carcinogenesis 2007
CALGB 80403/ECOG 1206: Survival
ECF-C
N=67
Mos
11.5
51
IC-C
N=71
FOLFOX-C
N=72
Total
N=210
95% c.i. Mos 95% c.i. Mos 95% c.i. Mos 95% c.i.
(8.1,12.5) 8.9 (6.2,13.1) 12.4 (8.8,13.9) 11.0 (8.8,12.3)
52
51
154
OS
median
# dead
PFS
median
5.9
# dead/pd 57
(4.5,8.3)
5.0
64
(3.9,6.0)
6.7
63
(5.5,7.4)
5.8
184
(5.1,6.8)
TTF median
5.5
#dead/pd/
off forAE 58
(3.9,7.2)
4.5
(3.6,5.6)
6.7
(4.8,7.2)
5.5
(4.5,5.9)
66
64
188
So what happened?
 ToGA worked and there is no real
signal from ECOG/CALGB trial
– Is this a failure of the “science?”
– Does HER 2 overexpression mean something
different than HER 1 (EGFR)?
Data from breast, colon and lung cancer don’t suggest
they are the same. Her 2 is a driver, but HER 1 is not in
most cases
– Am I reading too much into CALGB/ECOG
trial?
Authors of REAL3 and EXPAND hope so
Next Steps
 TOGA identified that for a small
subset of GE patients, trastuzamab
helps improve survival but not
miraculously
 We need to identify further targets,
optimization of identified targets and
how to combine targeted therapies
 How do we select?
HER family
 HER 1, 2, 4 transmembrane protein with
extracellular ligand binding domains and
intracellular tyrosine kinase domain
 HER 3 membrane bound with ligand
binding domain (neuregulin, for example)
but no kinase domain
– HER 3 heterodimerization may be key effector of HER2
activation
 Activation by ligand results in homo- or
heter-dimerizaiton and activation of PI3k
and/or MAP K pathways
www.cellsignal.com/reference/pathway/Erb
B_HER.html
www.cellsignal.com/reference/pathway/
ErbB_HER.html
Underlying etiology
 Find a driver of cell growth such as
Her 2
 One place to start is with the start of
the cancer
 Helicobacter Pylori
– Increasing evidence that CagA + H. Pylori
stimulates malignant transformation
– Some data exists on methods by which H.
Pylori stimulates this process
H Pylori
 CAG A + H. Pylori binds to CD 74
receptor
– This results in IL-8 production (other
cytokines) through NF-kappa B activation
– Induces CD 74 expression
– MEK inhibition may decrease the IL-8
production
Beswick EJ, et al J Immunol 2005;175;171-176
Sekiguchi H, et al Biosci Biotechnol Biochem. 2010;74:1018-24
What can CAG-A induce?
 It has been associated with
activation of
– EGFR (red because it appears to come
up with all the below)
MAP kinase
AKT
ERK
PI3 kinase
Nagy TA, et al J Infect Dis 2009; 199:641–5
Tabassam FH, et al Cellular Micro 2009:11, 70–82
Chen Y-C, et al W J Gastro 2006; 12: 5972-5977
Underlying etiology as a target
 Based on this evidence, EGFR should be a key
target
– Also, Ras mutations are rare (<2%)
 H. Pylori is not the only etiology of gastric cancer
and not the cause of esophageal cancer
– Likely more important in distal gastric cancer
– ECOG/CALGB studied esophagus and GE junction
 Is it possible EGFR inhibition is more relevant in H.
Pylori associated disease?
 Is it possible that if EGFR is an inciting event, by
the time we find the cancer, it is too late to block
EGFR and make a difference?
Target differences by location:
Not a new concept
 HER 2 overexpression by location
– GE junction: 24-32%
– Corpus/antrum: 10-18%
 HER 2 overexpression by
histology
– Diffuse: 2-7%
– Intestinal: 16-34%
Is it possible that activation of a kinase may
have different effects in different cells?
 Colon cancer and EGFR as an example:
– EGFR inhibitors don’t work (and in some settings
may harm) if Ras is mutated
Is PI3 kinase irrelevant in this setting?
Or is it that unbalanced stimulation of PI3K by EGFR
results in harm in some cases when Ras is also
activated?
 To butcher Orwell,
– “All pathways are equally relevant but some are
more equally relevant than others”, but which
pathway is more equally relevant varies
Every protein and its brother stimulates
PI3K and MAPkinase pathways
MET activation results in activation of beta catenin pathwasy,
PI3kinase pathway, STAT pathway and NOTCH pathway
MET appears to prolong Ras induction and Ras mediates MET
activation
C-MET
 I get to talk about it because they said I should talk
about all the proteins that interact with the HER
family and C-MET does
 Overexpressed in both gastric and esophageal
cancers
–
In esophageal cancer, expression increases in the pathway
from metaplasia to adenocarcinoma
 Can confer poor prognosis
 Preclinical studies show some cell lines
susceptible to MET inhibition
 C-MET can activate HER signaling through crosstalk with HER family receptors
Anderson MR, et al Clin Cancer Res 2006;12:5936-5943
Arkenau H-T J Cancer Res Clin Oncol 2009;135:855–866
C-MET inhibition
 C-Met inhibition has worked in
both gastric and esophageal cell
lines
– C-Met inhibition in esophageal cancer
In at least one cell line (Flo-1), MET
increased PI3K/AKT and PI3K inhibition
mimicked the effects of MET inhibition
That means that in Flo-1 cells, PI 3 kinase
mediates MET activity
Annotating the cell lines may help us
better understand what works and when
Watson GA, et al Neoplasia . 2006, 8:949 – 955
C-MET Inhibition
 Clinically, minimal data, but not
impressive
– Can the lab tell us why? One example:
Took C-MET addicted cell lines and made
them resistant to inhibition
– Displayed C-MET gene amplification
– Progressively developed Wt Kras amplification
– Cells became resistant to C-Met inhibition but
dependent on K-Ras
– Too bad we can’t inhibit Ras
Cepero V, et al Cancer Res 2010; 70; 7580–90.
Further complexity
 Src: a non-receptor tyrosine kinase
– Commonly activated in gastric cancer
– Induces activation of ERK and AKT
– In cell lines sensitive to Src inhibition, this results in
decreased phosphorylation of ERK and AKT and
increased apoptosis
 In a study of 16 gastric cancer cell lines, 14
with activated Src
– Dasatinib inhibited Src activated cell lines as long as
MET was not activated
– MET inhibition inhibited Src resistant cell lines with Src
activation
– Neither method inhibited cell lines without Src
activation
Okamoto W, et al. Mol Cancer Ther 2010;9:1188-1197
HER and MET
 Studies have now shown that activation of
HER family receptors confers resistance to
MET inhibition
– Activation of EGFR by providing ligand or by adding
mutated, constitutively activated EGFR led to AKT and
MAPK activation and resistance to MET inhibition
– HER 2/HER3 upregulation conferred secondary
resistance to CMET inhibition
– MET independent activation of HER family receptors
reactivates MAPK and AKT phosphorylation
Bachleitner-Hoffman, et al Mol Cancer Ther 2008;7:3499–508
Corso et al. Molecular Cancer 2010, 9:121
What can we learn from the limited
C-MET story I’ve shown?
 C-Met, like HER family has complex interactions
with other signaling pathways
 Upregulation of C-MET doesn’t always mean CMET sensitivity
 There are many mechanisms of resistance to CMET
–
–
–
–
Mutations in another pathway,
Upregulation of pathway proteins
Gene overexpression
Stimulation by ligand of another, complimentary (HER)
pathway
 In turn, C-MET activation can confer resistance to
inhibition of another pathway
Bringing it Back to HER
 Why is trastuzamab good, but no
miracle?
– Cross-talk between pathways is clearly
important
– Genetic variations in the interacting pathways
occur (eg AKT, PTEN, PI3kinase)
– Mutations in interacting pathways occur (Ras
2%, AKT 2%, PI3K 16%)
– All of the above can confer either primary or
secondary resistance
Hildebrandt MAT, et al J Clin Oncol 2009;27:857-871.
Barbi et al. Journal of Experimental & Clinical Cancer Research 2010, 29:32
Soung YH, et al Oncology 2006;70:285–289
 HER is at the top of the cascade,
why not go to the bottom?
www.cellsignal.com/reference/pathway/Erb
B_HER.html
www.cellsignal.com/reference/pathway/
ErbB_HER.html
The bottom (or close to it)
 Yoon, et al showed that different co-mutations led
to differential responsiveness of lung cancer lines
to MEK inhibition
 In HER activated, MET resistant cell lines, MAPK
and AKT inhibition were both required to inhibit
cell growth
 PI3K mutations confer resistance to MEK
inhibitors, especially in presence of Ras mutation
 And the list goes on
Yoon Y-K, et al MOLECULAR CARCINOGENESIS 49:353–362 (2010)
Bachleitner-Hoffman, et al Mol Cancer Ther 2008;7:3499–508
Wee S, et al Cancer Res 2009;69:4286-4293
Conclusions I
 Cancer is “driven” to grow
 The HER superfamily and its downstream
effectors are drivers of many
gastroesophageal cancers (and other
cancers too)
 Targeting this superfamily will likely yield
results
 But to truly win, we have to get smarter
Gettin’ Smarter
 Characterize our cell lines
 Determine which variables predict for
sensitivity to which inhibitors
 Test our patients for these variables
and treat based on these variables
 Determine mechanisms of resistance
so we are ready to adjust
 It’s time to split, not lump
– Challenge will be in trial design