matching evolving molecular diagnostics with novel...
TRANSCRIPT
Matching Evolving Molecular Diagnostics With Novel
Therapeutic Agents
Ana M. Gonzalez-Angulo, M.D. Associate Professor
Breast Medical Oncology Systems Biology
SOBO, Atlanta, GA 11/2012
Outline
• The basics • Challenges on molecular biomarker development
• New technologies and molecular diagnostics available
• How are we applying available molecular diagnostics to patient care today?
The basics
Evolution of the concept of breast cancer
Breast cancer = single disease with variable microscopic appearance
Breast cancer = single disease, variable microscopic appearance and variable
expression of estrogen/progesterone receptors
Breast cancer = at least 4 molecularly different diseases of the breast
Breast cancers = a collection of diseases with various combinations of deregulated
molecular pathways
Surgery (+/- chemotherapy)
Anti-estrogen therapy for estrogen receptor-
positive cancers
Various combinations of treatments based on
molecular type
Molecular pathway- directed therapies
Treatment strategy Definition of Disease
THERAPY DECISION-MAKING FOR EARLY BREAST CANCER
WHO CAN BE SPARED
THERAPY?
Prognostic markers needed
WHICH THERAPY WILL WORK BEST?
Predictive markers needed Modified from M. Piccart 2008
THERAPY DECISION-MAKING FOR EARLY BREAST CANCER
WHO CAN BE SPARED
THERAPY?
Prognostic markers needed
Identify patients at HIGH risk of recurrence and treat OR
Identify patients at LOW risk of recurrence and avoid the toxicity of adjuvant treatment
Modified from M. Piccart 2008
Examples of prognosis
Paik et al N Engl J Med 2004 Parker et al J Clin Oncol 2009 Bartlett et al Breast Cancer Res 2010
Oncotype DX RS
PAM50 ROR
Mammostrat
THERAPY DECISION-MAKING FOR EARLY BREAST CANCER
WHICH THERAPY WILL WORK BEST?
Predictive markers needed
Identify tumors with HIGH chance to response to an specific therapy
OR Identify tumors with LOW chance to
response to an specific therapy and discover new effective targets to treat them under clinical trials Identify patients with a HIGH chance to respond to an specific therapy OR
Identify patients with a LOW chance to respond to an specific therapy and find an alternative
PHARMACOGENOMICS Modified from M. Piccart 2008
THERAPY DECISION-MAKING FOR EARLY BREAST CANCER
WHICH THERAPY WILL WORK BEST?
Predictive markers needed
Identify tumors with HIGH chance to response to an specific therapy
OR Identify tumors with LOW chance to
response to an specific therapy and discover new effective targets to treat them under clinical trials Identify patients with a HIGH chance to respond to an specific therapy OR
Identify patients with a LOW chance to respond to an specific therapy and find an alternative
PHARMACOGENOMICS Modified from M. Piccart 2008
Examples of prediction of chemotherapy benefit
Paik et al. J Clin Oncol 2006 Albain et al Lancet Oncol 2010
Examples of prediction of chemotherapy response
Ignatiadis et al J Clin Oncol 2012
All
ER+/HER2- HER2+
ER-/HER2-
Different pathways are associated with pCR in different breast cancer
subtypes
Evaluation methods
Teutsch et al Genetics Med 2009 Modified from C. Sotiriou 2011
Gene/protein prognostic signatures
Add additional information to current clinico-pathological parameters for decision making for SOME patients
Oncotype DX RS
H/I + MGI
Mammaprint
GGI
PAM50
Mammostrat
Level I evidence to come!
Modified from C. Sotiriou 2011
Level I evidence is on the works!
PI: AM Gonzalez-Angulo
Tissues will be used to validate other genomic
signatures and compare them with the RS
Tumor burden may still matter • Prognostic signatures may not tell the complete
story!
Dowsett et al. J Clin Oncol 2010
>30% risk
HR-positive and HER2-negative breast cancer
Recurrence Score >25
Surgery: Number of positive nodes?
1-3 positive >4 positive
Adjuvant Chemotherapy
RANDOMIZATION Post-chemotherapy
Everolimus or Placebo
Everolimus for 1 year + appropriate endocrine therapy for 5 years
Placebo for 1 year + appropriate endocrine therapy for 5 years
Neoadjuvant chemotherapy:
Residual disease? Node-negative and tumor >2cm
Radiation therapy if indicated
>4 positive lymph nodes
Stratification factors: • Node negative • 1-3 positive nodes • >4 positive nodes Adjuvant • >4 positive nodes Neoadjuvant
SWOG S1207
PI: M Chavez-MacGregor
Stratification: •Endocrine therapy (Tamoxifen vs AIs) •Adjuvant chemotherapy
Pre and postmenopausal women with HR+ HER2– breast
cancer (≥ 4+ nodes Or N+ post-neoadjuvant)
Relapse-free after
2-3 yrs of adjuvant endocrine
therapy
N= 2010
Primary endpoint: DFS at 2 yr Secondary endpoints: OS, biomarkers, safety
Everolimus 10 mg/d for 2 yrs + AI or
Tamoxifen
Placebo for 2 yrs + AI or Tamoxifen
UNIRAD trial
PI: F. Andre
Molecular markers: challenges
Question on molecular markers
Since there are multiple targetable sites in a pathway, how can we ensure that the ones targeted and measured represent pathway activation and/or a predictor of therapy response?
To be considered: 1. Pathway drivers vs. passengers 2. Limitations in marker discovery 3. Pathway signatures should be more powerful than
single markers 4. Tumor evolution and marker changes
Trastuzamab Lapatinib Breast cancer HER2 amplification
Tamoxifen, AIs Breast cancer ER/PR expression
STRONG NEGATIVE PREDICTIVE VALUE <5% without the marker
HOWEVER POSITIVE PREDICTIVE VALUE LIMITED
30-60% with the marker
ENRICHES FOR RESPONDERS
Most effective targeted agents have efficient response predictor biomarker
RANDOM I ZAT I ON
Paclitaxel 175 mg/m2 q3w Lapatinib 1500 mg po QD
Paclitaxel 175 mg/m2 q3w Placebo po QD
Lapatinib for metastatic breast cancer
Di Leo et al, J Clin Oncol 2008
P + L (n = 52)
P (n = 39)
Median, mos 8.1 5.8
HR (95% CI) 0.57 (0.34, 0.93)
P value 0.011
ErbB2-positive ErbB2-negative P + L
(n = 199) P
(n = 202)
Median, mos 5.8 5.3
HR (95% CI) 1.04 (0.83, 1.30)
P value 0.747
Total 75 events (83%) Total 299 events (75%)
0
20
40
60
80
100
0 3 6 9 12 15 18 21 24 27 30 Time, months
Cumulat
ive
prog
ress
ion-
free
, %
P + Lapatinib P + placebo
0
20
40
60
80
100
0 3 6 9 12 15 18 21 24 27 30 Time, months
P + Lapatinib P + placebo
TTP by ErbB2 status
Di Leo et al, J Clin Oncol 2008
Sources of IHC Testing Variation
Preanalytic Time to fixation Method of tissue processing Time of fixation Type of fixation
Analytic Assay validation Equipment calibration Use of standardized laboratory procedures Training and competency assessment of staff Type of antigen retrieval Test reagents Use of standardized control materials Use of automated laboratory methods
Postanalytic Interpretation criteria Use of image analysis Reporting elements Quality assurance procedures Laboratory accreditation Proficiency testing Pathologist competency assessment
Limitations in marker discovery
Courtesy L. Middleton 2010
Pathway signatures should be more powerful than single markers
• Driver abnormalities can be identified by the integration of multidimensional data sets
Berns et al, Cancer Cell 2007
P=0.127 P=0.052 P=0.007
Molecular evolution and marker changes
Gonzalez-Angulo et al Mol Cancer Ther 2011
Molecular evolution and marker changes
Gonzalez-Angulo et al Mol Cancer Ther 2011
New technologies and molecular diagnostics available
Comprehensive molecular portraits of human breast tumours
TCGA Network Nature 2012
Significantly mutated genes and correlations with genomic and clinical features.
Comprehensive molecular portraits of human breast tumours
TCGA Network Nature 2012
Somatic alterations observed in human breast cancer samples
Stephens et al Nature 2009
Interpretation of data
Hamptom et al Genome Res 2009
A sequence-level map of 157 chromosomal breakpoints in a single
human breast cancer cell line
Genomic data in absence of cell biology, may not
be informative
NGS available in the clinic
This does not mean that this molecules will have a
clinical effect
NGS available in the clinic
How are we applying available molecular diagnostics to patient
care today?
NOT standard of care practice, but under controlled clinical trials
Tandem, 2-step, phase II marker evaluation design
Some responses
Complete step 2 of the study
Few or no response
Start a second 2-step study for marker evaluation Treat marker-positive cases only
Few or no response Some responses
Complete step 2 of the study Failed predictor!
No pressure for patient selection!
Potentially useful marker that defines Drug sensitive patient population
Pusztai et al Clin Cancer Res 2007
Treat all comers, with early stopping rule
Dose and schedule for targeted therapies
• What is the activating mechanism for cancer growth? • Oncogenic addiction (Mutations/Amplifications)
• Transient but complete inhibition (High Cmax) • No oncogenic addiction (WT)
• Sustained and complete target inhibition (AUC) • Secondary resistance
• Give the drugs in sequence
• This may be even more relevant in combinations
PRESENTED BY: Ana M. Gonzalez-Angulo
Where are we now? • Our medical practice is based on standards of care (EBM)
• EBM: best approach for the average populations, not for specific individuals
• Application of systems biology to personalized cancer
therapy- Breast cancer as a model • Molecular profiling technologies to tailor medical care
• Challenges:
• Identifying and validating molecular markers • Molecular crosstalk and bypass mechanisms • High failure rate of molecular targeted therapeutics
• It is critically important to understand the pathways and
networks to target as well as of the homeostatic loops induced by the interventions
It is much more complex
Next Generation Sequencing Reveals Co-Activating Events in the MAPK and PI3K/AKT
Pathways in Metastatic TNBC
Informed Consent on USON IRB-approved protocol N=1 CLIA
Life SOLiD 4
• Multiple subtypes of mTNBC with heterogeneous biology
• Growth and Survival Pathways of Interest in mTNBC include:
DNA repair defects IL-6/jak-2/stat-3 PI3K
VEGF Androgen receptor Src
EGFR FGFR ERK
O'Shaughnessy, et al. SABCS 2011
PI3K
KRAS GTP
RAS GDP
RTK
BRAF
S6K Cell Survival Cell Proliferation
MEK
ERK
AKT
mTOR
INPP4B
PTEN
NF1
IQGAP3
FBXW7 = mutation
= mRNA up
= mRNA dn
= DEL
= AMP
SUMMARY PI3K and MAPK Pathways
ARAF
ERAS GTP
O'Shaughnessy, et al. SABCS 2011
Inhibition of MEK results in compensatory increase in PI3K/Akt,
particularly in cells with loss of PTEN
Mirzoeva et al. Clin. Cancer Res 2008 Saal et al. Nat. Genetics 2007
Modified from. Arteaga at SABCS 2009
Combined inhibition of PI3K and MEK is effective in all basal-like preclinical
breast cancer models
Modified from. Arteaga at SABCS 2009 Hoeflich et al. Clin Cancer Res 2009
Not all tumors with the same aberration have the same response
PRESENTED BY: Ana M. Gonzalez-Angulo
All tumors have RAS/RAF mutations
BKM120 + GSK1120212
Bedard P et al. ASCO 2012
Challenges in combination of targeted therapies
• How to choose a dose escalation schema • Toxicity issues:
• Individual drug toxicities • Overlapping toxicities • Unexpected toxicities • How we choose which drug to reduce?
• If we can only completely inhibit one pathway, which pathway is the dominant one?
PRESENTED BY: Ana M. Gonzalez-Angulo
Why are PD studies important?
• They can provide answers with different consequences in drug development:
• Lack of clinical activity:
• Pathway not inhibited: Dosing/schedule? • Pathway inhibited: Explore underlying biology
• Presence of clinical activity: • Pathway not inhibited: Mechanism of action? • Pathway inhibited: Is this a good predictor
PRESENTED BY: Ana M. Gonzalez-Angulo
Conclusions • Understanding cancer as a molecular disease is leading to novel therapeutic
approaches, more individualize
• Today,pognostic and predictive signatures can help in treatment-decision for specific groups of patients
• Appropriate biomarker selection is increasingly important for the integration of novel agents in the treatment of cancer
• Further translational studies are needed to clarify the different mechanisms of pathway activation, thus it can be targeted appropriately
• New clinical trial designs are needed to rapidly evaluate the hundreds of
targeted therapeutics and potential biomarkers that are under preclinical evaluation
• Clinical trials to evaluate the role pathway inhibitors at different levels and combination strategies are ongoing and should have Strong preclinical background extensive correlative components to be able to decipher the best use of these drugs according to the molecular aberrations of the tumors
• Contribute to clinical trials !!!
Acknowledgements Mentorship • G.N. Hortobagyi • G.B. Mills • F. Meric-Bernstam
Gonzalez-Angulo’s Lab • S. Liu • B. Wang • C. Phan • H. Chen • E. Tarco • N. Parinyanitikul • J. Sohn • A. Trape Meric-Berstam’s Lab • A. Akcakanat • G. Singh
Collaborators MDACC Systems Biology • K. Hale • J. Mendelsohn Transcriptional Profiling • L. Pusztai • W.F. Symmans Tumor Bank • A. Sahin BMO • L. Hsu
Surgical Oncology • E. Mittendorf
Bioinformatics • K. Coombes • Y. Qi • Z. Ju • W. Liu Biostatistics • D. Berry • K. Do • X. Lei
T and H&N • G. Blumenschein Phase I • Razelle Kurzrock
Collaborators Outside MDA • C. Perou, L. Carey • I. Krop • R. Bernards, H. Horlings • A. Lluch, J. Ferrer • C. Arteaga • J. Baselga • J. Tabernero, J. Rodon • J. Gray • M. Ellis • C. Hudis, N. Rosen • C. Sotiriou • P. Lorusso • AL. Borresen-Dale • F. Andre • M. Pollak
Funding By • NIH • MDACC Physician-Scientist Start up Funds • Komen for the Cure • BCRF • Texas Fed of Business and Professional Women • Commonwealth Foundation for Cancer Research • AACR SU2C Dream Team • ACS •PI of ISTs with Novartis, BMS, GSK, Abraxis, Roche Dx, Genomic Health Inc, Merck, Genentech, Bayer, EMD. • Lab MTAs with NIH, Merck, Exelixis, Novartis, Xcovery, EMD Serono, Genentech, Bayer
