Identifying Tumors
Expressing Predictive
Markers: Lessons Learned
From HER2
Dennis J. Slamon, MD, PhD
TRIO Chairman
Chief, Division of Hematology/Oncology
David Geffen School of Medicine at UCLA
Los Angeles, California
Faculty Disclosure
Dennis J. Slamon, MD, PhD, Speakers Bureau:
Genentech/Roche, GSK, sanofi-aventis
Advisory Board: Novartis
Molecular Diversity of
Human Cancers:
Biologic and Therapeutic
Implications
BRCA1
HER2
Paradigm Changes from
Human Breast Cancers
Human Breast Cancer Is Highly Heterogeneous
STAGE
Lymph. infiltrate
invasive
In situ
low
Well-
Margins
Nuclear Grade
Differentiation
Poorlyhigh
infiltrating
“pushing”
“single-file”
Can we decipher new molecular genetic information for
these complex and variable tumors and establish a new
classification with real therapeutic impact.
THE PAST
The “One-Size-Fits-All”
Approach to Cancer
Cell Type and Phenotype
TDLU
K18
K14
CALGB 9344: Overall
Survival
Henderson, et al. J Clin Oncol. 2003;21:976-83.
9
Breast Cancer Subtypes are
associated with disease outcome
Sørlie et. al. PNAS 2003
CURRENT THERAPEUTIC BREAST CANCER
SUBTYPES
15-18%
20-25%
60-65%
Triple-Negative Breast Cancers: Some
Potential Therapeutic Targets
Cetuximab
EGFR
Tyrosine
Kinase
MET
tyrosine
kinase
MAP Kinase Pathway
MAPK inhibitors;
NOTCH inhibitors
AntiAngiogenesis
MET mab
Akt Pathway
Transcriptional Control
Cell
Cycle
PARP inhibitors
DNA
Repair
pathways
Bevacizumab
Cell Death
After Cleator S et al. Lancet
Oncol. 2006:8:235-244
Can We Do Better?
The Hope - Clinical Translation of
Biologically Relevant Molecular
Information Should Lead to More
Effective and Less Toxic
Therapeutic Approaches
CURRENT TRANSLATIONAL
RESEARCH PROCESS
Hypothesis
Generation
BASIC SCIENCE
LABORATORIES
Tissue
Specimens
TRANS
CLINICAL
TEAMS:
Protocol
Development
Specimen/Sample
BASIC SCIENCE
LABORATORIES
The HER2 Alteration
Southern
Northern
Western
IHC
Slamon et al. Science 198
Slamon et al, Science 1987
HER-2 Oncogene
Amplification
Breast Cancer
HER-2 Oncoprotein
Overexpression
Shortened Survival
Median Survival from First Diagnosis
HER-2 overexpressing
HER-2 normal
3 yrs
6 - 7 yrs
Target Validation - A
Biologic Effects of HER-2/neu
Amplification/Overexpression in Human Breast
Cancer Cells
DNA Synthesis
Cell Growth
HER2Breast
Cancer Cell
Lines
HER-2
Transfection
HER2+
Breast Cancer
Cell Lines
Growth in
Soft Agar
Tumorigenicity
Metastatic
Potential
E2 Response,
Tam Resist.
Target Validation - B
Preclinical Impact of Trastuzumab
on Tumor Growth
Effect of Trastuzumab Treatment on HER2+ Breast Cancer Xenografts
Tumor volume (mm3)
2000
Control
Trastuzumab
1500
1000
Trastuzumab
withdrawn
500
0
0
10
20
30
40
Treatment day
Pietras et al. Oncogene. 1998;17:2235.
50
60
70
Trastuzumab in Combination
with Chemotherapy
Objective - Combination Compared to Chemotherapy Alone
 Primary
– Time to disease progression (REC)
– Safety
 Secondary
–
–
–
–
–
Overall response rates
Durations of response
Time to treatment failure
1-year survival
Quality of life
Summary: Phase III Clinical Trial Comparing Best Available
Chemotherapy to
Chemotherapy+Trastuzumab
Enrolled 469 pts
H +CT 235 pts
CT 234 pts
RR
49% (^53%)
32%
Resp Duration
9.3M (^59%)
5.9 M
TTP
7.6M (^68%)
4.6M
The HER2 Alteration
Southern
Northern
Western
IHC
Slamon et al. Science 1987,1989
Disease-Free Survival
B-31
100
N9831
100
AC TH
87%
90
AC T
AC TH
90
85%
87%
AC T
78%
80
80
86%
74%
%
70
66%
70
N Events
60 AC T
807 90
AC TH 808
HR=0.55, 2P=0.0005
50 51
N Events
60 AC T
872 171
AC TH 864
-9
HR=0.45,
2P=1x10
83
50
0
1
2
3
4
68%
5
0
1
Years From Randomization
2
3
4
5
Lessons from the HER2 Story
1.) Target Identification
2.) Target Validation
3.) Preclinical Confirmation
4.) Determintion of Potential Usage
Preclinically
5.) Clinical Translation - Proof of Concept
6.) Clinical Optimization
Other Lessons Learned:
What we are learning about already established
agents
The META-Analysis
How Did The Current Chapter
Start ?
Attempts to explain the differential
prognosis of HER2 positive breast
cancers
The HER-2 Gene: encodes a 185kd
protein that is a member of the type I
receptor tyrosine kinase family which
also contains EGFR, HER-3 and HER-4
Functions When Altered:
1.) Growth and proliferation - increased
2.) Differentiation - decreased
3.) Cell survival - increased
4.) Motility - increased
5.) Neoangiogenesis - increased
6.) Reduced dependency on estrogen and insensitivity to hormonal
blockade
HER-2 neg
MA-5 TRIAL
HER-2 pos
Pritchard, NEJM 354:2103, 2006
Disease Free Survival
95% CI
HER2 positive
HER2 negative
Study
HR
NSABP B11
0.60
0.96
0.84
1.02
0.65
1.35
0.83
1.22
0.75
0.79
0.52
0.91
0.44 - 0.82
0.75 - 1.23
0.65 - 1.08
0.86 - 1.20
0.34 - 1.27
0.93 - 1.97
0.46 - 1.49
0.91 - 1.64
0.53 - 1.06
0.60 - 1.05
0.34 - 0.80
0.71 - 1.17
0.90
0.71
1.00
0.82 - 0.98
p = 0.01
0.61 - 0.83
0.90 - 1.11
p < 0.0001
p = 1.0
NSABP B15
Brussels
Milan
DBCCG-89-D
NCIC MA-5
Total
Overall
heterogeneity c25 = 5.3, p = 0.38
heterogeneity c25 = 7.6, p = 0.18
anthra better
0.4
0.6
Test for interaction chi2 = 13.7 p < 0.001
0.9
non anthra better
1
2
5
A. Gennari,
JNCI 2007
Overall Survival
HR
Study
NSABP B11
NSABP B15
GUN 3
Milan
DBCG-89-D
NCIC MA-5
Total
Overall
0.66
0.90
0.82
1.07
0.85
1.64
0.61
1.26
0.73
0.82
0.65
1.06
0.91
0.73
1.03
0.47 CI
95%
0.92
0.69 0.63 1.06
1.18
0.88 0.27
1.30
2.69
0.85 3.150.32
1.16
0.89 1.790.50
1.05
0.59 1.130.42
1.01
0.80 1.40
0.83 - 1.00
0.62 0.85
0.92 p = 0.391.16
heterogeneity c25 = 5.2,
heterogeneity c25 = 5.5, p = 0.36
HER2 positive
HER2 negative
anthra better
non anthra better
p = 0.056
p < 0.0001
p = 0.86
0.4
0.6
Test for interaction chi2 = 12.0, p < 0.001
0.9
1
2
5
A. Gennari,
JNCI 2007
The Topoisomerase IIa Gene:
encodes an enzyme which is critical
in DNA replication and function
including RNA transcription
Functions:
1.) Resolves topological problems in DNA
2.) Is critical in RNA transcription from DNA
3.) Makes transient protein-bridged DNA breaks on one or both DNA strands
during replication
4. Plays critical roles in segregation, condensation and superhelicity
The Topo IIa protein is a
major target of the
anthracyclines
Can We Do Even Better?
The Hope - Further Clinical
Translation of Biologically
Relevant Molecular Information
Should Lead to Even More
Effective and Less Toxic
Therapeutic Approaches
CURRENT TRANSLATIONAL
RESEARCH PROCESS
Hypothesis
Generation
BASIC SCIENCE
LABORATORIES
Tissue
Specimens
TRANS
CLINICAL
TEAMS:
Protocol
Development
Specimen/Sample
BASIC SCIENCE
LABORATORIES
Clinical Outcome in Primary
Papillary Serous Carcinoma
Disease Free Survival
Overall Survival
≈ 60% recur within 2 years
≈ 20% mortality within 2 years
≈ 75% recur within 3 years
≈ 40% mortality within 3 years
Primary Papillary Serous
Complete
Censored
1.0
0.9
Cumulative Proportion Surviving
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0
365
730
1095
Surv iv al Time
uncensored: 83 ( 83.00%)
censored: 17 ( 17.00%)
uncensored: 57 ( 55.34%)
censored: 46 ( 44.66%)
1460
1825
Goals
Identify molecular subtypes of ovarian tumors that
may have clinical and biological relevance for
disease initiation and progression
Utilize these data to generate and test therapeutic
hypotheses
Build on the work done in other programs
Cedars-Sinai/UCLA Ovarian
Cohort
 225 ovarian samples have been received from Dr. Beth Karlan of
Cedar Sinai, profiled and imported into Rosetta analysis software
– Samples collected between 1989 and 2005
– RNA quality measured using Agilent BioAnalyzer
– RNA Integrity Number (RIN) average = 9.16
 All samples were profiled using Agilent Human 1A V2 chip
– Reference is an equal mixture of the first 106 ovarian samples
profiled
 Detailed clinical outcome is available on 90% of the samples
 UCLA has completed FISH analysis and/or Northerns for a number
of genes including HER2, EGFR, Periostin (POSTN, PN)
UCLA/Cedar Sinai Ovarian Tumor
Study: Papillary Serous
Characteristic
No. of patients
(%)
(N=132)
Characteristic
No. of patients
(%)
(N=132)
Recurrence
Age
< 50 yr
31 (23.5)
≤ 12 months
55 (41.7)
≥ 50 yr
98 (74.2)
> 12 months
46 (34.8)
Missing
3 (2.3)
Progressive/Refractory
NED
Missing
Stage
I
II
4 (3.1)
17 (12.8)
5 (3.9)
4 (3)
5 (3.8)
Tissue Status
III
95 (72.0)
Primary
IV
21 (15.9)
Recurrence
106 (80.3)
20 (15.2)
Locally advanced
1 (.75)
Interval
1 (0.75)
Missing
6 (4.5)
Missing
5 (3.8)
NED: No evidence of disease
Hierarchical Cluster of Ovarian
Samples across 6165 Genes
Normal samples
(n=14) show a very
similar pattern of
gene expression
Unsupervised
clustering does not
group remaining
samples into clear
subtypes
Refine Analysis to Discover Ovarian
Subtypes
Unsupervised hierarchical clustering clearly defines only
a normal & “normal-like” subtype
Clinical outcome does not define subgroups
– ANOVA based on overall survival finds 0 differentially
expressed genes (DEG)
Consider other markers to distinguish ovarian subgroups
– Periostin (POSTN, PN) & TGFβ Induced (TGFβI)
– Hormone receptor markers: AR, PGR, ER
– CA125 (MUC16)
Refine Analysis to Discover Ovarian
Subtypes
Unsupervised hierarchical clustering clearly defines only
a normal & “normal-like” subtype
Clinical outcome does not define subgroups
– ANOVA based on overall survival finds 0 differentially
expressed genes (DEG)
Consider other markers to distinguish ovarian subgroups
– Periostin (POSTN, PN) & TGFβ Induced (TGFβI)
– Hormone receptor markers: AR, PGR, ER
– CA125 (MUC16)
225 Ovarian Samples Clustered across 2830
Genes identifies three major subtypes
Normal
POSTN
ER
NORMAL
AR
PR
POSTN
TGFβI
CA125
ER
Clinical Outcome in Primary
Papillary Serous Carcinoma
Disease Free Survival
Overall Survival
≈ 60% recur within 2 years
≈ 20% mortality within 2 years
≈ 75% recur within 3 years
≈ 40% mortality within 3 years
Primary Papillary Serous
Complete
Censored
1.0
0.9
Cumulative Proportion Surviving
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0
365
730
1095
Surv iv al Time
uncensored: 83 ( 83.00%)
censored: 17 ( 17.00%)
uncensored: 57 ( 55.34%)
censored: 46 ( 44.66%)
1460
1825
POSTN Signature Related to Clinical
Outcome in Primary Ovarian
Samples
Disease Free Survival
Overall Survival
Primary Ovarian Samples: POSTN in Primary Ovarian Samples
Complete
Censored
Overall Survival: POSTN in Primary Ovarian Samples
Complete
Censored
1.0
1.0
p=0.008
p=0.03
0.9
0.9
n=113
0.8
0.8
0.7
0.7
0.6
0.6
n=31
0.5
0.5
n=112
0.4
0.4
0.3
n=29
0.3
0.2
Probability of Remaining Re
0.1
0.0
0
365
730
Time
1095
1460
1825
Group 0.
Group 1.
Cumulative Proportion Surviving
0.2
0.1
0.0
-0.1
0
365
730
Time
1095
1460
1825
Group 0.
Group 1.
NORMAL
AR
PR
POSTN
TGFβI
CA125
ER
POSTN Signature Related to Clinical
Outcome in Primary Ovarian Samples
Disease Free Survival
Overall Survival
Primary Ovarian Samples: POSTN in Primary Ovarian Samples
Complete
Censored
Overall Survival: POSTN in Primary Ovarian Samples
Complete
Censored
1.0
1.0
p=0.008
p=0.03
0.9
0.9
n=113
0.8
0.8
0.7
0.7
0.6
0.6
n=31
0.5
0.5
n=112
0.4
0.4
0.3
n=29
0.3
0.2
Probability of Remaining Re
0.1
0.0
0
365
730
Time
1095
1460
1825
Group 0.
Group 1.
Cumulative Proportion Surviving
0.2
0.1
0.0
-0.1
0
365
730
Time
1095
1460
1825
Group 0.
Group 1.
Challenges to new and/or combined
use of targeted therapeutics
Identifying the appropriate patient population
Do we simply integrate new targeted therapies with
established regimens? Advantages/Problems
Is broader target specificity better than more narrow
targeting?
What are the most rational targeted combinations to test
clinically?
Can we determine the best likely combinations pre-clinically
before going into the clinic? Challenges - predictive value of
models