Molecularly Targeted Therapy in Lung Cancer: Hype, Hope, Myths and Reality

Molecularly Targeted Therapy in
Lung Cancer: Hype, Hope,
Myths and Reality
Martin J. Edelman, MD
University of Maryland Greenebaum
Cancer Center
Advanced NSCLC: The State of
the Art
What is Targeted Therapy?
• If we use the analogy of pesticides: empiric
therapy would be “Raid” while targeted therapy is
the “Roach Hotel.”
Dr. David Gandara
• A “smart” bomb versus a “cluster” bomb.
Dr. Nevin Murray
Targeted Therapy: A definition
• Drugs targeted at pathways, processes and
physiology which are uniquely disrupted in cancer
cells:
–
–
–
–
Receptors
Genes
Angiogenesis
Tumor pH
• Get real, these pathways etc. are not so distinct.
Six Essential Alterations
in Cell Physiology in Malignancy
Self-sufficiency in
growth signals
Evading
apoptosis
Hanahan & Weinberg,
Cell 100:57 (2000)
Insensitivity to
anti-growth signals
Targets for classical drugs?
Targets for novel drugs?
Sustained
angiogenesis
Tissue invasion
& metastasis
Limitless replicative
potential
Myth #1: Targeted therapy is new
(or what were we thinking?)
•
•
•
•
DNA is not a bad target.
Tubulin is a very good target.
Specific receptor targeting: tamoxifen etc.
Sometimes you design for one target and hit
another: estramustine.
Survival by B-Tubulin III Phenotype
In Taxane-treated NSCLC
1,0
0,9
p=0.02
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0,0
0
2
4
6
8
10
Months after diagnosis
12
14
16
Dumontet,
2002
Better Identify and Utilize the
Drugs We Already Have: GILT
R
A
N
D
O
M
I
Z
E
Standard Arm
Docetaxel/Cisplatin
Genomic Arm
Beta tubulin ERCC1 Regimen
Doc/CDDP
+
Doc/Gem
+
Gem/CDDP
+
+
Gem/CPT-11
New Anti-tubulins: Epothilones
Computer model of
beta-tubulin mutations
Taxane
Epothilone
Giannakakou et al. Proc. Natl. Acad. Sci. USA
2000, 97, 2904
Myth #2: Imatinib Mesylate is
the “Proof of Principle” for these
Drugs
• Imatinib Mesylate targets the bcr-abl TK very
specifically.
• Bcr-abl is the root cause of CML, essentially a
“monogenetic disease”
Probability of Relapse
Imatinib Mesylate: The Exception that
Proves the Rule…
Time to Relapse in Patients
with Myeloid or Lymphoid Blast Crisis
Who Had a Response to STI571
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Myeloid (n=21)
Lymphoid (n=14)
0
100
200
300
400
• Orange arrows indicate patients still enrolled in the study and in remission at the
time of the last follow-up
• White arrows indicate the day on which patients were removed from the study
NEJM 344: 1038, 2001
Targeted Therapy in the Common
Solid Tumors: The Reality
But what about Trastuzumab?
• Degree of benefit is
relatively modest.
• Population is enriched
by Herceptest.
• Drug would have
<10% RR if entire
population treated
The EGFR (ErbB) family and ligands
EGF
TGF
Amphiregulin
-cellulin
HB-EGF
Epiregulin
Tyrosine kinase
domain
Heregulins
NRG2
NRG3
Heregulins
-cellulin
100
44
36
48
100
82
59
79
100
33
24
28
ErbB-1
Her1
EGFR
ErbB-2
Her2
neu
ErbB-3
Her3
Cysteine-rich
domains
C-terminus
ErbB-4
Her4
Why the Difference?
• Most solid tumors have complex genetics,
not one or two hits but 20+. The more
advanced the tumor, the greater the
heterogeneity.
• Molecular heterogeneity.
• Hitting one narrow target is not likely to be
that beneficial.
Tumors Progressively Make More Angiogenesis
Stimulators
bFGF
bFGF
bFGF
bFGF
VEGF
VEGF
VEGF
PDGF
PDGF
IL-8
Relf et al., Cancer Research, 57:953,
1997
How to hit the target
• If you know the target, and there is only one target
you can be very specific.
• If you don’t really know or it’s a really big target,
a larger weapon may be needed.
But all is not lost…
• Return to the fundamental assumption.
• Targeted therapy works when you can
identify and validate the target.
– Need to enrich the population for the target:
Herceptin
– May need to hit more than one target
– Importance of trial design
Leveraging your opponents weight, or how
targeted therapy can work with other
treatments and toss the opponent out of the
ring
Arachidonic Acid Metabolism
Cell membrane phospholipids
Sphingomyelin
Neutral sphingomyelinase
Arachidonic Acid
COX 1,2
5-HPETE
12-HPETE 15-HPETE
Celecoxib
5LO
12HETE
5HETE
Zileuton
LTA4
LTB4
LTC4
LTD4
LTE4
Ceramide
15HETE
Prostaglandins
CALGB 300203:Gemcitabine/ Carboplatin
+ Eicosanoid Modulators
PD
Stage IIIB (pleural effusion),
IV NSCLC
PS 0-1
Adequate organ fcn
Carboplatin AUC =5.5
Gemcitabine 1000 mg/m2
Zileuton 600 mgpo qid
SD, PR
CR
Eicosanoid modulator
until progression
Carboplatin AUC =5
Gemcitabine 1000 mg/m2
Celecoxib 400 mg po bid
Carboplatin AUC =5.5
Gemcitabine 1000 mg/m2
Zileuton 600 mgpo qid
Celecoxib 400 mg po bid
MJ Edelman, PI
Off study
Correlates: IHC, CYFRA
VEGF levels
Issues in Trial Design
• How to screen drugs
• Identify logical targets
• Identify whether target acquired and
neutralized.
• Identify the population with the target.
• Was the target important.
• Beware collateral damage.
The Phase I Trial
• Traditionally, based upon
the idea that good
medicine tastes bad.
– A little is good, more is
better.
– Clearly false
• If you know the target,
you can find the dose.
– Demonstrate target
inhibition
– Downstream effects
– Pharmacokinetic parameters
Molecular Target Measurements
• Enzyme activity measurements
• e.g., DT-diaphorase, P450’s
• Gene mutation status
• e.g., Ras, p53
• Protein levels
• e.g., Thymidylate synthase
• mRNA levels
• e.g., Microarray
• Other
• e.g., Phosphorylation status of a protein
Clinical Trial Design With
Biologic Endpoints
• Evaluate for target effect as active concentration is
approached
• Expand cohort when any biologic effect seen
– reproducibility of effect
– importance of well defined confidence interval
• Escalate dose
– until maximal expected effect is seen
– until maximal effect occurs in maximal fraction of patients
• Additional steps to confirm
– effect is maximal
– rate of effect is maximal
Phase I of Targeted Agent: Use of
Enzyme Inhibition to Determine Dose
The Oncologist, 2002 7:401-409
Phase I Trial of a Targeted Agent: Use of
Target Inhibition to Determine Dose
Trial Design: The Value of the
Phase II Study
• Early optimism led to the assumption that one
could go directly from Phase I to III.
• Unless a home run is assured, not a good idea.
Decisions made in haste are repented at leisure
• Phase II designs with “go and no go” endpoints.
• New Phase II designs
– Randomized Phase II
– Enrichment designs
– Discontinuation designs
Powering the Phase II Trial
• Response rate is a crude intermediate
marker. The various Phase II approaches
are adaptable to any endpoint.
• Power for TTP, survival or a molecular
marker.
• But whatever you do, identify an endpoint.
False Positive Trials
• Patient self-selection
• Physician selection: incentives for entry
• Assessment of responses: the power of
wishful thinking
• And of course, inadvertent enrichment for
the target
Phase III Trials: The Problem of
Molecular Heterogeneity
Impact on Clinical Trials Outcome
• Histologic diagnosis remains a key
eligibility criterion.
• Tumors with indistinguishable
histology demonstrated quite different
responses to therapy.
• Do solid tumors possess different
“destinies” based on their molecular
profiles?
Anaplastic Oligodendroglioma
• Proportion of genetic subtypes differ in
different cohorts of patients.
– in recurrent disease: 90% 1p LOH
– in newly dx’d disease: 60% 1p LOH
• Genetic subtype is age dependent.
– 1p LOH more common in young pts
– older pts predominate in trials
0.8
0.6
0.2
0.4
1p LOH
Neither 1p LOH
nor TP53
0
Survival Probability
1.0
Molecular Subtypes of Anaplastic Oligodendrogliomas: Implication for Patient
Management at Diagnosis
Y. Ino et al. Clin Cancer Res 7:839,2001
0
50
100
Months
150
200
False Negative Trials
• Overestimation of a therapeutic effect due to
enrichment of phase II studies for a treatment
sensitive subtype.
• Dilution of a beneficial effect in responding
patients by large numbers of nonresponding
patients.
• Reversal of a beneficial effect in responders by
negative effect in nonresponders.
Identify Specific Subsets
• The promise of
genomic and
proteomic technology.
• This enriches the
population.
• We already do this:
a
b
– PML
– CML
– Breast cancer
Pharmacogenomics
Drug metabolizing
enzyme
Frequency of variant poor Drug
metabolism phenotype
CYP 2D6
6.8% Sweden
1% China
Codeine
CYP 2 C9
3% England
Warfarin
CYP 2C19
2.7% white USA
14.6% China
18% Japan
Omeprazole
DPD
1% of the population
5FU
Targeted Therapy: The Future
• Modern biology has identified a host of new
potential targets for cancer therapy
• Drugs interacting with these targets are available.
• The benefit of these agents is dependant upon the
criticality of the target. More than one target may
need to be inhibited.
• New agents may “tip the balance” when combined
with chemotherapy, radiation.
Targeted Therapy: The Future
(cont’d)
• The design and careful assessment of new agents
in Phase I and II trials will result in better
understanding of the potential population and
magnitude of benefit for any particular agent.
• Phase II trials should guide the decision for Phase
III.
• Phase III trials with a good chance of success can
then be accomplished.