Clinical Trial Preparation for Molecularly Based Agents

1
Clinical Trial Preparation for Molecularly Based
Agents

• Edward A. Sausville, M.D., Ph.D
• on behalf of Dr. Louise Grochow
• Chief, Investigational Drug Branch
• Cancer Therapy Evaluation Program
• National Cancer Institute

T
C
EP
THEORY TO THERAPY
2
Changing the paradigm targeted therapy
development
Agent Selection
Murine tumors Xenograft models
Credentialed targets molecular models
Priority
Log cell kill
Growth inhibition
Trial design
Empirical
Hypothesis driven
Dose endpoint
Toxicity
Molecular effect When to measure?
Endpoint eval
HP, clin. lab
Complex assays
Eligibililty
Any solid tumor
Presence of target
3
Changing the paradigm Phase II
Ph II goal
Tumor shrinkage (cure)
Tumor stabilization (eliminate progression)
Ph II metric When?
Anatomic imaging 8 weeks
Functional probes Establish time points
Dose finding
More is better
Molecular effect
Combinations
Empiric
Pathways and preclinical proof of principle
Patient selection
Histology
Molecular pathology
4
What experts will you need to work with?

• Basic, translational and clinical colleagues
• Preclinical experts in model design, toxicology
  and activity
• Clinical trials experts
• Statisticians
• Experts in molecular target assessment
• Pathologists
• Imagers
• Interventional radiologists

5
So you have a good idea Now what?

• Characterize the target
• Screen libraries against the target OR
• Synthesize based on in silico drug design
• Select lead compound based on pharmacology
• Screen for activity in engineered lines
• Validate in animal models mimicking human disease
• Develop reporters/probes to assess target effects
• Design a clinical trial

6
Clinical Trial Design dose finding (phase I)
studies

• Establish disposition and safety
• Establish dose that produces target effect
• (Target Effect Dose)
• Establish relationship between
• dose/schedule and effect
• Agent may be non-toxic (endostatin)
• (Expensive) agent may have effects at doses that
  arent toxic
• Dose response relationship may not be monotonic
  (interferon)

7
Response ()
Relative Concentration (LOG scale)
8
Possible Trial Designs

• PK/target concentration-guided
• Accelerated Titration Design
• Minimize patients treated at inactive doses
• Incorporate biological target effect
• Other designs incorporating biological effect
• O6BG AGT suppression
• single dose before planned tumor resection
• PS 341 proteasome inhibition
• circulating normal PBM
• EGFR inhibitors assess target presence
• bcr/abl tyrosine kinase inhibitor response

9
O6BG Rationale for Development

• Alkylator-resistant tumors often have increased
  intracellular alkylguanine-DNA alkyltransferase
  (AGT), DNA repair enzyme
• O6BG depletes AGT
• When AGT lt 10 fmol, activity of DNA alkylating
  agents that form adducts at the O6 guanine
  position increases
• O6BG and BCNU more active in vivo than BCNU alone

Friedman et al JCO 163570, 1998
10
O6-BG Phase 1 Trial Design in Malignant Glioma

• Undetectable AGT (lt 10 fmol/mg protein) occurs in
  20 in absence of O6-BG
• O6-BG administered 18 hr before craniotomy
• BCNU-induces interstrand cross-links by 18 hrs
• Up to 13 patients entered at each dose
• At any time, if AGT detectable in 3 patients,
  dose was escalated
• Biologic endpoint undetectable AGT in gt 11 of 13
  patients

Friedman et al JCO 16 3570, 1998
11
Glioma AGT Activity After O6-BG
AGT Undetectable
Dose
No. of
(mg)
Patients
0
9
0
40
3
0
60
3
0
80
13
8
100
11
11
Friedman et al JCO 163570, 1998
12
O6-Benzylguanine Phase I Trial

• Suppression of AGT activity in peripheral blood
  mononuclear cells (PBMC) did not predict
  suppression of AGT activity in tumor tissue in
  phase I studies
• PBMC 36 mg/m2
• Tumor 120 mg/m2
• Immunoreactive AGT in PBMC a poor surrogate

Spiro TP, et al. Clin Cancer Res. 72318, 2001
13
Clinical Trial Design withBiologic 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

14
Support for Early Clinical Trials

• CTEP support (agent supply)
• CTEP support for infrastructure
• Cooperative agreements (Dose finding)
• Contracts (Phase II)
• Other support for infrastructure
• Grant support
• Quick trials (R21)
• Clinical study section
• Cancer Center Core Grants
• General Clinical Research Centers

15
Applying to CTEP for Agents

• Letter of Intent
• Form available on CTEP web site
  http//ctep.info.nih.gov
• Review criteria
• Strong scientific hypothesis
• Supporting preliminary data
• Adequate patient accrual
• Innovative correlative studies
• Ability to meet regulatory requirements
• Not duplicative
• Agent available
• Industry sponsor concurs

16
Preparing a Protocol

• LOI approved after reviewed by CTEPs Protocol
  Review Committee ( extramural investigators if
  needed)
• Address critique in LOI approval letter
• Use available templates (supplied by CTEP)
• Include relevant details (or references) for
  correlative studies
• Submit electronically to CTEP PIO
• Review by CTEP PRC

17
What do you need to measure drug effect on target?

• Characterized assay or probe to report effect on
  target relevant to therapeutic agent
• Quality assurance calibration assays,
  interfering processes, sample handling
• Reproducible
• Sensitivity
• Specificity
• Economically and practically feasible
• Validate in engineered model

18
Desirable Information for Design of Dose -
Finding Trial

• Relevant models
• Optimal schedule
• Optimal dose
• Concentration required for effect on target
• change in target associated with efficacy
• Variability in dose/target and target/efficacy
  relationships
• Estimate of fraction of tumors that will have
  effect
• Normal target values, variability
• Target effect on tumor vs other tissues (eg PBM)
• Does patients tumor have relevant target?

19
CORRELATION BETWEEN 20S PROTEASOMEINHIBITORY
POTENCY GROWTH INHIBITIONFOR 13 DIPEPTIDE
BORONIC ACIDS
Correlation r20.92
Mean GI50 (nM)
Ki (nM)
Adams et al, Cancer Res 592615, 1999
20
EFFECT OF PS-341ON PC-3 TUMOR GROWTH IN MICE
Tumor Volume ( Vehicle)
Week
Adams et al, Cancer Res 592615, 1999
21
EFFECT OF PS-341ON 20S PROTEASOME ACTIVITY
Mouse WBC
PC-3
20S Activity ( Vehicle)
20S Activity ( Vehicle)
Adams et al, Cancer Res 592615, 1999
22
Ex Vivo Proteasome Activity1 Hour Post Treatment
20S Activity
1.96 mg/m2
PS-341 (Log dose, mg/m2)
23
MS-275 HDAC INHIBITOR NSC 706995

• Collaboration between NCI and Nihon Schering
• Very differential activity in NCI screen
• High oral bioavailability
• Range finding tox suggests GI/Marrow DLT
• on qd x 14 schedule

24
EFFECT OF HDACIs ON HISTONE ACETYLATION
PC-3M Prostate Cancer Cells
Peripheral Blood Lymphocytes
Acetylated H3

Acetylated H4
A. Control B. TSA, 0.3?M, 8hr C. MS-275,
0.3?M, 24hr D. MS-275, 1.0?M, 24hr
1. Control 2. MS-275, 0.3?M, 2hr 3. MS-275,
1.0?M, 2hr 4. MS-275, 0.3?M, 8hr 5. MS-275,
1.0?M, 8hr
6. MS-275, 0.3?M, 24hr 7. MS-275, 1.0?M,
24hr 8. TSA, 0.3?M, 2hr 9. TSA, 0.3?M,
8hr 10. TSA, 0.3?M, 24hr
25
CELL CYCLE PHASE ANALYSISOF MS-275 ON PROSTATE
CANCER CELLS
PC-3M, 24h Treatment

26
EFFECT OF HDACIs ON p21waf1 EXPRESSION
Control

TSA 0.3 ?M 8 h
MS-275 1 ?M 24 h
MCF-7 Breast Carcinoma
Du145 Prostate Carcinoma
27
REAL-TIME PCR OF p21 cDNA IN PC-3MAFTER HDI
TREATMENT

?Rn
Cycle Number
28
Options for Preliminary Efficacy (phase II)
Trials

• Conventional and/or molecular target eligibility?
• Standard phase 2 design targeting RR, survival,
  PFS, clinical benefit
• Standard with biologic endpoint
• Standard targeting non-progressor rate
• Ratio of TTP for new and previous therapy
• Tumor growth before and after treatment with each
  patient as her own control
• Multi-arm randomized selection designs
• Randomized discontinuation design

29
Clinical Trial Design Combination Studies

• Combinations of novel agents, novel agents with
  cytotoxics
• Preclinical models or plausible molecular
  hypothesis
• Biologic target effect designs
• proof of principle
• Dose escalation schemes

30
Tumor Tissue Studies in Clinical Trials

• Difficulty obtaining tumor tissue
• Difficulty obtaining multiple samples
• Which time points?
• Tumor heterogeneity results vary depending upon
  tumor vs normal vs necrotic cells
• Methodological issues sample handling,
  sensitivity, reproducibility, complexity,
  availability
• For newer targets, limited information on normal
  ranges, variability, magnitude of desired effect
• 192 biopsies in 107 pts 88 success in obtaining
  paired biopsies

Dowlati, Haaga et al.. Clin Cancer Res 72971,
2001
31
Setting priorities for clinical trials

• Credentialed target
• Evidence that agent effects target
• Evidence that target effect correlates with
  useful therapeutic effect
• Reliable way to measure presence of target
• Probes that permit assessment of target effect in
  clinical trial

32
Selecting and Evaluating New Agents for Cancer
Treatment

• Test strategies for selecting agents for trials
• Test hypothesis of molecular diagnosis
• Evaluate novel clinical trial designs
• Target effect is necessary but not necessarily
  sufficient (pathway crosstalk, downstream events,
  other proliferative advantages
• If proposed target is not accurately
  characterized, a useful agent might be discarded
• This is much simpler if the agent is non-toxic,
  cheap and highly selective in its effects on a
  critical target
• This is much harder if the agent is promiscuous
  in its targets, toxic, and the molecular
  pathology of the disease is complex

33
Acknowledgements

• O6BG H. Friedman, S. Gerson, H. Spiro, E.
  Dolan, J. Pluda
• Biopsy data J. Haaga, A. Dowlati
• PS341 J. Adams
• MS275 J. Trepel