1. Re-Engineering Early Phase Cancer Drug
Development: Decreasing the Time from
Novel Target to Novel Therapeutic
James H. Doroshow, M.D.
Deputy Director for Clinical & Translational Research
National Cancer Institute, NIH
16th Annual Drug Discovery Symposium
Robert H. Lurie Comprehensive Cancer Center &
Center for Molecular Innovation & Drug Discovery
Northwestern University
Chicago, IL
October 12, 2011

2. Most Drugs Fail in Late Stages of Development:
Particularly in Oncology
Rates of success for compounds entering first
in man that progress to subsequent phase
All indications • 70% of oncology drugs that
enter Phase 2 fail to enter
Phase 3
• 59% of oncology drugs that
enter Phase 3 fail
• Late stage failure leads to
enormous risk
• Failure is primarily due to
lack of efficacy>toxicity
Kola & Landis; Nature Reviews Drug Discovery 2004

3. Standard Drug Development Pipeline:
Re-envisioned
Hypothesis Clinical
Generation Candidate Development Commercialization
Drugs Assays Trials $1200 MM
Cumulative
Investment
Risk
Target $500-600 MM
Validation
Target/
Molecule Assay
Lead Preclinical
Discovery Development Optimization Development Phase Phase Phase Regis- Global Global
I II III tration Launch Optimization
Lead
Generation
$200-300 MM
$20-60 MM Risk
GOAL:
Time: 12-15 Years
Time: 6-8 Years

4. Therapeutics Discovered / Developed by NCI from
Preclinical Stage Approved by FDA Past Decade
Year Agents Role of NCI Mechanism of Support
2011 18F-NaFluoride Produced agent; filed first-ever NDA from an DCTD Frederick personnel;
(PET Bone Scans) NIH Institute: developed distribution system extramural contracts; DCTD
after NDA approved; bone scans absent 99Tc regulatory support
2010 Eribulin Natural product discovery; screening; clinical DCTD/DTP Frederick labs;
formulation; efficacy testing; clinical candidate analytical, formulation, PK,
selection; first-in-human trials by NCI toxicology contracts; RO1
grant; U01 grant
2009 Pralatrexate RAID project; NCI produced GMP bulk drug DCTD contract resources for
production of GMP quality bulk
drug
2009 Romidepsin (Depsipeptide) Developed safe human dosing schedule in large DCTD/DTP pharmacology and
animals; PK and Tox; produced drug for clinical toxicology and drug
trials; conducted first-in-human trials in NIH CC production contracts;
Frederick animal facilities
2004 Cetuximab Produced first lots for imaging; chimeric Frederick Biologics Contracts;
clones; NCI Drug Development Group Grantee Coop. Drug Discovery Grant
2004 5-Azacytidine Pre-clinical molecular pharmacology; produced DTP Contracts; Frederick Labs;
pre-clinical and clinical drug supply; conducted U01 Grants
pivotal trial
2003 Bortezomib Extensive analog screening; MOA and PD DCTD/DTP Frederick labs;
studies; PK & Tox; clinical formulation formulation, PK, Tox contracts
2000 Temozolomide Scale up synthesis and clinical formulation DCTD/DTP bulk drug and
formulation contracts

5. NCI’s Approach to Discovering &
Developing More Anticancer Drugs
• Provide better discovery and development tools
ü Public availability of screening data from approved &
investigational agents
ü Targeted combinations—in vitro and in vivo
• Facilitate academic discovery per se
• Provide structured resources to transition from
R01 target discovery to the clinic

6. US FDA Approved Anticancer Agents
Activity in the NCI 60 panel: Potency in µM
Name NSC # Mean GI50 Mean TGI50 Mean LC50
Signaling agents
Bortezomib 681239 0.00051 0.0063 3.6
Dasatinib 732517 0.33 8.9 51
Erlotinib 718781 5.5 59 >90
Everolimus 733504 0.095 14 56
Gefitinib 715055 3.2 19 49
Imatinib 743414 15 43 81
Lapatinib 745750 2.9 20 61
Nilotinib 747599 2.9 13 49
Sorafenib 747971 1.9 6.0 30
Sunitinib 750690 2.2 9.6 31
Temsirolimus 683864 0.038 51 >100
Romidepsin 630176 0.00025 0.0081 0.038
Vorinostat 701852 0.94 17 70
Molec. Cancer Ther. 9: 1451-1460, 2010

7. Plated Drug Sets: Dilution Series
http://dtp.cancer.gov/branches/dscb/oncology_drugset_explanation.html

8. Transformation of the NCI Therapeutics Pipeline
Cancer SPORE RO1/PO1 Biotech &
Small Pharma
Centers
Roadmap
Intramural
Imaging/IDG
RAID DDG
CBC Created
The NCI Experimental Therapeutics (NExT)
Pipeline:
Target discovery through early stage clinical trials
Exploratory Screening/ Phase 0 / Early Phase
Screen Designed Lead Candidate Clinical II Trials
I Trials
Development Synthesis Development Seeking Candidate
Drug Discovery Early Development
8

9. Where Did We Need to Go?
Rapid translation of discoveries into public health benefits
Created NCI Experimental Therapeutics Program (NExT):
Unified Discovery & Development utilizing NCI Frederick
as a Critical Resource
A single pipeline for all therapeutic development resources: One
Pipeline, Many Points of Entry
NExT Program Coop Grps
INCLUDES
Targets • Investigational drugs and biologics
Therapeutics
• Investigational imaging agents and theranostics
• Academic & Biotech & Pharma projects
• Includes Phase 0, I and II Programs

10. Goals of the NExT Program
• Develop
treatments
for
unmet
medical
needs
(e.g,
rare
cancers
and
pediatric
tumors)
• Provide
resources
for
natural
product
development
and
the
development
of
high
risk
targets
• Move
discoveries
from
TCGA
into
drug
discovery
•
Support
development
of
biological
agents
• Develop
qualified
PD
and
predicEve
molecular
markers
for
the
clinic
• Focus
on
discovery
and
development
projects
from
academic
invesEgators
and
small
biotechs

11. NCI Chemical Biology Consortium (CBC)
• Mission: Dramatically increase flow of early stage drug
candidates into NCI therapeutics pipeline
• Vision:
• Develop integrated network of chemists, biologists,
and molecular oncologists, with synthetic chemistry
support
ü Active management by NCI and external
advisory boards
ü Unify discovery with NCI pre-clinical and
clinical development
ü Linked to other NCI initiatives; NCI Intramural
chemistry integral partner
• Focus on unmet needs in therapeutics: “undruggable”
targets, under-represented malignancies
• Enable a clear, robust pipeline all the way from target
discovery through clinical trials for academic, small
biotech, and pharma investigators
FRONT END: NCI’s Experimental Therapeutics Platform 11

12. Chemical Biology Consortium Vision
Why is CBC different?
• Builds on >50 yrs of NCI experience
in cancer drug development
• Not intended to replicate Pharma NCI
• CBC members will submit own
projects and take on those of other
investigators
• Focus on bringing academic targets
Discovery Risk
and molecules to patients
• Will not shy away from difficult Biotech
targets
• Longer time horizon
• NCI committed to supporting CBC Big
projects from inception through
proof-of-concept, PD-driven clinical Pharma
trials if milestones achieved: Only
NCI could do this
• Inclusive involvement of CBC
members in shared projects
Market Risk
developed in parallel across
consortium

13. NCI Chemical Biology Consortium (CBC)

14. Multiple Entry Points into the NExT
Exploratory Screen Screening/
Development Designed Lead Development Candidate Clinical
Synthesis Seeking Candidate
Parallel Optimal Efficacy in
Target
Primary HTS medicinal potency/ pivotal in vivo
identification
chemistry selectivity models
Model Development and Small Animal Early Stage Later Stage Assay
Target Validation Imaging Center Pharmacodynamics Validation

15. Therapeutics Discovery & Development Support
Provided by NCI (NExT)
• Medicinal
chemistry,
HTS,
lead
op7miza7on
• Enhanced
synthesis
of
small
molecules
and
pep7des
• Scale-­‐up
produc7on
of
small
molecules,
biologicals,
&
imaging
agents
• Isola7on
and
purifica7on
of
naturally
occurring
substances
• Development
of
early
stage,
clinical
pharmacodynamic
assays
• Exploratory
toxicology
studies
and
pharmacokine7c
evalua7on
• PK/PD/efficacy/ADME
studies
(bioanaly7cal
method
development)
• Development
of
suitable
formula7ons
• Range-­‐finding
ini7al
toxicology
and
IND-­‐directed
toxicology
• Product
development
planning
and
advice
in
IND
prepara7on
• Later-­‐stage
preclinical
development
of
monoclonal
an7bodies,
recombinant
proteins,
and
gene
therapy
agents
• Manufacture
of
drug
supplies
• Analy7cal
methods
development
for
bulk
material;
formula7on
• CLIA-­‐grade
clinical
assay
development
for
later
trials
• Produc7on
of
clinical
dosage
forms
• Stability
tes7ng
of
clinical
dosage
forms
• Regulatory
support
and
early
phase
trials

16. NCI Experimental Therapeutics
How Does An Extramural Investigator
Access NCI’s Drug Discovery and
Development Resources?

17. NExT Application Process
Extramural scientists may propose targets, screens, or
molecules for entry into the NExT pipeline; 3 receipt dates
per year
https://dctd.cancer.gov/nextapp or
https://dctd.cancer.gov/nextregistration

18. NExT Pipeline: Oversight and Decision Support
Anticipate ~20-30 projects in the pipeline per year

19. Prioritization Process Used To Ascertain Which
Compounds To Move Forward?
• This selection is based on the following criteria.
– Scientific Merit Scoring:
1 = Exceptional
– Feasibility 3 = Excellent
– NCI Mission 6 = Satisfactory
– Novelty 9 = Poor
– Clinical Need
• A Stage Gate evaluation process to benchmark the
progress and priority of projects within the portfolio
• This evaluation process is also to provide guidance
about the priority utilization of the capacity – based
resources provided by NCI

20. 265 NExT Applications Received in Cycles 1-8
Source of NExT Applications

21. LDHA: Therapeutic Target in Cancer
• The proto-oncogene c-myc can drive glutamine as well as glucose
metabolism. In cancer, c-myc deregulation can result in the added
uptake of glucose and its conversion to lactate, thereby contributing to
the “Warburg Effect”.
• ChIP sequencing confirmed that Lactate Dehydrogenase A (LDHA), an
enzyme that converts lactate to pyruvate, is a direct downstream target
of Myc.
• Knockdown of LDHA decreased colony formation and reduced the
growth of tumors in breast and lung cancer xenografts.
• Japanese families that completely lack LDHA are otherwise normal
except for exertional myopathy.
• FX11 is a selective, small molecule, active site LDHA inhibitor identified
from a malarial LDH screen that provides proof-of-concept for targeting
cancer metabolism in human lymphoma and pancreatic cancer models.

22. FX11 Treatment Leads to Regression of Tumors
in Lymphoma and Pancreatic Xenograft Models
Lymphoma Pancreas
FX11

23. LDHA : Next Steps
Screen
development and
Hit to Lead Lead Optimization Candidate Seeking
high-throughput
screening
Primary uHTS NCGC: hit validation/med Co-crystallization with HTS “hits”
chemistry
NIH Chemical
Genomics Center Secondary biochemical
and cell-based screens
Dang Lab
FX11 Lead Compound
Ki 4 uM
ü HTS screen to identify new scaffolds
ü Co-crystallization with FX11
ü Optimize SAR for lead compound FX11, increase potency and improve solubility

24. Portfolio Stratified by Agent Class
Natural
Product
Imaging
5%
3%
Biologic
35%
Small
Molecule
57%
Projects that are closed or awaiting resourcing are not included

25. Interim Project Evaluation
Projects
Closed
by
NExT
Bayer
Healthcare
(Including
Legacy
Projects
from
RAID)
In-House Target Validation
Nat.
Rev.
Drug
Discov.
10:
643-­‐644,
2011
3%
33%
33%
64%
67%
§ Results Not Replicated
§ Results Not Replicated § Reproducible
§ Did Not Meet Milestones
§ Not applicable

26. ASSAYS: Proof of Mechanism in Early Trials
• Demonstrate drug action on intended tumor target (proof of
mechanism) in a human malignancy early in development
– evaluate hypotheses surrounding mechanism of action per se
– evidence of target modulation in the clinic assists decision to move agent forward,
or not . . .
– evaluate relationship of drug schedule and systemic exposure to target effects;
examine relevance of marker chosen to represent target modulation
– prior to expectation of efficacy
• Potential to investigate molecular effects of the agent in non-
malignant tissues
– relevance of ‘surrogate’ tissues
– genetic toxicology
• NOT: predictive of clinical benefit
– only later stage (larger) trials can define relevance of target modulation to tumor
growth inhibition
Clin. Cancer Res. 14: 3658-3663, 2008

27. Developing the ‘Right’ Assay Tools
for Early Stage Proof of Mechanism Studies

28. Clinical Qualification of PD Assay:
“Humanizing” Preclinical Models
“Clinical Readiness” of PD Assay:
Therapeutically relevant in preclinical model: Replicate the clinical setting in
which the assay will be practiced:
• Clinical procedures for sample acquisition and handling
• Storage requirements and transferability
• Time frames achievable in clinical setting
• Inter- and intratumoral variability
• Stability of baseline
• Minimum doses required for target effect; does target inhibition correlate with
altered tumor growth or toxicity in vivo
• Suitability of surrogate tissues
• Actually works when you do a “DRY RUN”
Now You Can Start!

29. Standard 18 gauge Bx Cryobiopsy: Freeze
Cryobiopsy: Excise Excisional Biopsy

30. Comparing Effect of Four Tumor Harvest
Methods on phosphoAKT Levels
phospho-AKT
+C C011 Untreated Jurkats
Resection After CO2 SAC
Resection After CO2 SAC
(Cell Signaling #9271)
Cryobiopsy + Anesthesia
Cryobiopsy + Anesthesia
Resection + Anesthesia
Resection + Anesthesia
FNA + Anesthesia
FNA + Anesthesia
200
140
pAKT Settings
100
Min 20.0
80
Max 75.0
60 60 kDa 1 min exp.
50
40
30 368
β-Actin
Actin Settings
Min 20.0
Max 3000.0
30 sec exp.
Separated on an 8% Tris-Gly Gel

31. Indenoisoquinolines
• Unique, non-camptothecin
Topo I inhibitors; chemically stable
• Prolonged Topo I-Drug-DNA complex
formation
• Unique patterns of DNA cleavage
• Not substrates for ABCG2 efflux pump
• Produce dose- and time-dependent
DNA double strand cleavage
demonstrable as phosphorylation
of the H2AX histone
• Low cross-resistance with camptothecin
analogs (irinotecan; topotecan)
• Discovered by Yves Pommier (NCI intra-
mural program); developed by DCTD
• FIH Randomized NCI Phase I trial of
NSC 724998 vs 725776
Develop comprehensive PD package
for proof of mechanism evaluation
PRIOR to first-in-human studies

32. Development of a Validated ELISA for
Topo I Quantitation: Indenoisoquinoline First-in-Human Trial
Topo I ELISA Correlation of Topo I mRNA (Array)
to Protein (ELISA) in NCI 60 Cell Panel
12
R2 = 0.6915
HRP
Y
Y
Y
Topo I (ng/mL/µ g protein)
Y
Y 10
Y
8
Top1 6
Y
4
2
Y = MAH IgM MAb C21 (BD) capture Ab
Y = RAH pAb Ab28432 (AbCam) primary Ab 0
Y-HRP= GAR-HRP-XSA secondary Ab 50 100 150 200 250 300 350 400
(XSA =extra serum absorbed)
Topo I mRNA (average microarray intensity)
Comparison of Topo I Protein by Western Blot Effect of Topotecan on
to Protein (ELISA) Topo I Activity in vitro (ELISA)
100
ELISA
WB
75
Relative Topo I level
50
25
A375 Melanoma
0
HCT-116 HT-29 H322M SR SKMEL28 A498 HL-60*
Molec. Cancer Ther. 8: 1878-1884, 2009

33. Topoisomerase I Inhibition and Correlation with Efficacy
In Vivo Using Topo I ELISA
Topoisomerase I Levels in Xenograft Extracts
AAXR2-18, YKR2-39, YPR2-2, AAYR2-17
Treated with Topotecan or Vehicle Control
Effect of NSC 724998 on Topo I Levels in
A375 Xenografts
Topo 1 (ng/ml/µg ptn)
Vehicle Control
4h Topotecan
(15 MG/KG) treated +
NSC 724998 (mg/kg)
Vehicle Controls
Solid red line = Avg vehicle control
Dashed red line = Avg ± 1 and 2 SD
Black line = Dose Response

34. Pharmacodynamic Assay Development for Proof of Mechanism Early Phase
Trials: A Resource Intensive Enterprise
REQUIRES
• Develop “FDA-quality” assays for target status, downstream
effects, and toxicity markers
• Relate drug effect markers to tumor effect and drug
exposure in preclinical models of Phase 0/1 target lesions
• Develop procedures for biopsies of tumor and normal tissues
using clinical instruments that provide samples available for
biomarker assay development that mimic what is available in
the clinic
• Formalize biopsy, specimen handling, and analytical
assays as SOPs
• Implement PD assay SOPs in clinical arena of operations
• Conduct laboratory analysis of clinical specimens
• Laboratory QA/QC; qualify outside labs for testing, and
prepare qualifying sample sets and analytes to disseminate
technology

35. Current Stages of Therapeutics Development: Trials
N=30 N=300 N=3000 FDA
Phase I Phase II Phase III

36. Match the Technology to the Task

37. Improving the Impact of Early Clinical Trials
Target Assessment
Preliminary estimate accuracy of measurement
of drug effect on target
Downstream molecular interactions
Early read on efficacy
Understanding toxicity
Supporting assays N=10-60 N=200-300 N=600-800 FDA
in model systems
Target Development
Qualify Assays Target Validation
Initial Proof of Mechanism Comparative benefit of target inhibition
Optimize Association of MOA with efficacy
“Phase 0”
Assay Minimum data to define safety

38. Target Inhibition as the Endpoint of a Phase II Trial:
Proof of Concept Study of Oral Topotecan in Advanced
Solid Neoplasms Expressing HIF-1α
NCI-05-C-0186: Giovanni Melillo, MD PI
• Eligibility: HIF-1α +ve solid tumors of any histology (>10% of
tumor cells by IHC)
• Treatment: Oral chronic topotecan (1.2 mg/m2 PO daily x 5 days
x 2 wks q28 days)
• Primary endpoint: Inhibition of HIF-1α expression in tumor
• Schema:
Biopsy
PET Biopsy
DCE-MRI PET PET CT
CT DCE-MRI DCE-MRI
D1 D8 D29 D36
Cycle 1 Cycle 2
PD endpoints:
• IHC (MVD, Glut-1)
• mRNA expression (HIF-1 target genes,
VEGF, PGK-1, CAIX)
• serum/plasma markers (VEGF, osteopontin)
• CEP (circulating endothelial precursor cells)

39. Pilot Study of Oral Topotecan in
Advanced Solid Neoplasms Expressing HIF-1α
ü Accrual: 16 patients
• 12 evaluable: 1 melanoma, 1 bladder, 1 breast, 2 ovarian ca., 1 SCLC,
1 bladder, 1 H/N, 4 CRC [PRs in SCLC, Ovarian cancer]
• 4 not evaluable: 1 ASPS, 1 adrenal, 1 colon, 1 pancreas
ü Toxicities: myelosuppression, diarrhea (first 2 pts., at 1.6 mg/m2),
well tolerated at 1.2 mg/m2
HIF-1α staining in patient #4 (breast cancer)
Baseline Biopsy After 2 Cycles of Topotecan
Clin. Cancer Res. 17: 5123-5131, 2011

40. mRNA Expression of HIF-1α Downstream Targets
After Topotecan
160
VEGF Baseline
140
Post-treatment
Percentage expression 120
100
80
60
40
20
0
Pt 1 Pt 2 Pt 3 Pt 5 Pt 10 Pt 15
250
GLUT-1 Baseline
Percentage expression
200 Post-treatment
150
100
50
0
Pt 1 Pt 2 Pt 3 Pt 5 Pt 10 Pt 15
Clin. Cancer Res. 17: 5123-5131, 2011

41. Evolution of the
Oncologic Drug Development Paradigm
1991 2011
• Develop estimates of response • Perform flexible evaluations of
rates and characterize side effects therapeutic activity and toxicity over a
in a non-randomized fashion broad range of continuous endpoints
• For minimally pretreated patients • For a wide variety of: acceptable
subjects, prior treatments, extent and
with measurable disease after evaluability of disease and
“standard” therapy performance status
• In a specific tumor type • In a wide range of genotypically,
• With a drug (that has a non- rather than histologically, specific
specific mechanism of action) diseases, under conditions that allow
assessment of the biochemical effect
of an intervention in tumor and normal
tissue
• With one or more molecules that
produce specific effects on defined
targets (as well as possible off-target
interactions)
• Using the minimum necessary data
set to assure the safety and efficacy of
a novel therapy

42. DCTD Division of Cancer Treatment
and Diagnosis
Accelerating Cancer Diagnosis and Drug Development
v Developmental v DCTD
Therapeutics
Jason Cristofaro
Jerry Collins
Joe Tomaszewski Barbara Mrochowski
Melinda Hollingshead
Ralph Parchment v CTEP
Robert Kinders Jamie Zweibel
Tom Pfister
Jeff Abrams
Myrtle Davis
Bev Teicher
Shivaani Kummar v Cancer Imaging
Paula Jacobs
v Center for Cancer
Research
Yves Pommier v Cancer Diagnosis
Lee Helman Barbara Conley
Bob Wiltrout Mickey Williams
William Bonner