Transcript Slide 1

How to Design High Impact Trials
to Indentify Biomarkers
Janet Dancey, MD
Ontario Institute for Cancer Research
NCIC Clinical Trials Groups
2nd Quebec Conference on Therapeutic Resistance
Montreal, November 5-6th 2010
Potential Conflict of Interest
• Dr. Janet E. Dancey
– None
Objectives
• Types of biomarker studies
• Uses in clinical trials
• Methods and design issues
3
Why do biomarker studies?
• Biology
◦
◦
◦
Understand cancer
Identify new targets for therapeutics
Identify new markers of diagnosis, prognosis, prediction, monitoring
• Diagnosis
◦
◦
To identify site of origin of an undifferentiated tumour,
To identify second primary from metastases
• Prognosticate
◦
To predict outcome (risk of toxicity, relapse, progression)
• Predication
◦
To predict benefit/risk (or lack) from a specific treatment
• Monitor
◦
Identify cancer early, monitor response/progression
4
Why do biomarker studies?
• To understand cancer biology
• To improve treatments
• To change medical practice
5
Why is doing biomarker studies so difficult?
• Cancer models are not patients and people are not
laboratory models
• ….and its not just biology
Things you don’t hear in the lab
By the way….
• My family has been in-bred for
generations
• Those cancer cells in my flank
had been in culture for decades
No visit, treatment, biopsy, imaging
today, please….
• I’m not well
• the insurance won’t cover it
• the REB says you can’t
6
Sometimes it’s where the needle went
Intratumoral heterogeneity of carbonic anhydrase IX (CAIX)
Effect of distributional heterogeneity
on the analysis of tumor hypoxia
based on carbonic anhydrase IX
VV Iakovlev, M Pintilie, A Morrison, et al
Laboratory Investigation (2007) 87, 1206–1217
a) Immunoperoxidase staining for CAIX in a
single tissue section. Analysis of the entire
section gave a value of 10.8% CAIX labeling.
The circles limit the analysis to 0.6 mm
simulated tissue microarray (TMA) cores, and
show a wide range in CAIX (for publication
purpose only, the image was digitally enhanced
to better visualize CAIX areas).
7
Sometimes it’s the timing
pSer473-Akt antibody in human GI tumors and HT-29 colon cancer
xenografts measured by immunohistochemical staining.
surgically specimens
biopsy specimens
Baker, A. F. et al. Clin Cancer Res 2005;11:4338-4340
A, patient tumor samples. 1 and 2 are two surgically resected specimens and 3 and 4 are two biopsy specimens.
B, HT-29 human tumor xenografts excised from scid mice and kept at room temperature for the times shown. Each section
also includes in the upper right-hand quadrant an on-slide control of HT-29 colon cancer cells stained for pSer473-Akt.
Copyright ©2005 American Association for Cancer Research
8
Sometimes it’s how we measure it
9
Why are successful biomarker studies
uncommon?
• Biological heterogeneity
◦ Cellular, tumour, patient
• Assay variability
◦ Within assay, between assays
• Specimen variability
• Effect size
A lot of “noise” that blur marker and outcome correlation and
validation
10
‘Validation’
Feinstein
• “Validation is one of those words — like health,
normal, probability, and disease — that is constantly
used and seldom defined.
• We can ... simply say that, in data analysis, validation
consists of efforts made to confirm the accuracy,
precision, or effectiveness of the results.”
Feinstein, A. R. Multivariable Analysis: An Introduction (Yale University
Press, New Haven, 1996).
11
Biomarker Validation
Laboratory
• Biomarker – marker of biology;
◦ Scientific validation
◦ Technology/analytical validation
Years
• Assay – method/means of measurement;
• Test - clinical context
◦ Clinical validation/qualification
• Clinical utility
◦ Value of using the test versus alternatives
Clinical Uptake
Multistep, multi-year, interative process requiring multidisciplinary collaboration
12
Trial Designs and Biomarkers
Trial Phase
Purpose
Biomarkers
Modifications
0
Define dose
Select agents
Target modulation
PK
Normal Volunteers
Pre-surgical
I Metastatic
Dose/schedule
Target Modulation
PK
Toxicity
Activity
Expanded cohorts
to evaluate target ,
toxicity or screen
activity
II Metastatic
Activity
Predictive markers
Randomized
III Metastatic
Clinical benefit
Predictive markers
Subset analyses
III Adjuvant
Clinical benefit
Predictive
Prognostic
Subset analyses
Type of marker depends on trial phase
13
Phase 1 Trials: Considerations
• Primary goal: To identify an appropriate
dose/schedule for further evaluation
Small
patient
• Design principles:
numbers
◦ Maximize safety
◦ Minimize patients treated at biologically inactive doses
◦ Optimize efficiency
• Study population:
◦ Patients for whom no standard therapy
Heterogenous
Refractory
Tumours
Expect target modulation but not anti-tumour activity
14
Where/when do biomarkers play a role?
Target Versus Toxic Effects
Probability of Effect
1.0
Off Target Toxicity
Target Effect in Tumour
Target Toxicity
Target Toxicity
Dose/Concentration/Exposure
15
Biomarker Studies in Phase 1 Trials
• MGMT activity after O6-benzylguanine
 Friedman H et al J Clin Oncol 16:3570-5, 1998; Spiro et al. Cancer Res
59:2402-10 1999; Dolan et al Clin Cancer Res 8:2519-23, 2002
• 20S proteosome inhibition after bortezomib
 Lightcap E et al. Clin Chem 46:673-683, 2000; Adams J, Oncologist 1:
9-16, 2002;
• DCE-MRI after PTK787/valatanib
 Galbraith S et al NMR in Biomed 15:132-142, 2002; Morgan, B. et al. J
Clin Oncol; 21:3955-3964 2003;
• S6K inhibition after everolimus
 Tanaka C et al J Clin Oncol 26:1596-1602, 2008
• PARP Inhibition after ABT-888
 Kinders RJ, et al. Clin Cancer Res. 2008 Nov 1;14(21):687785
• ERK Inhibition after PLX-4032
 Puzanov, K. L. J Clin Oncol 27:15s, 2009 (suppl; abstr 9021)
16
PLX4032, a V600EBRAF kinase inhibitor: correlation of
activity with PK and PD in a phase I trial.
Puzanov, K. L. J Clin Oncol 27:15s, 2009 (suppl; abstr 9021)
Patients
pERK
PRE
pERK
KI67
PRE
KI67
PK
µM*h
Imaging
4
range
50-100,
median
60;
range
10-40,
median
11
range
20-60%,
median
45%;
range
5-25%,
median
12.5%
mean
AUC024h ~
126
µM*h
PD (4)
500 1000
PR (1)
↓ PET (2)
2
70
5-fold
35-fold
Target
4-fold
2
30 -50%
3-5%
10-fold
Pathway
Tumor
17
Phase I Predictive Markers
Target
Drug
Test
Phase I ORR (%)
PARP
Olaparib (AZD2281; KU- BRCA1/2
0059436)
9/21 (44%) Ovary,
breast, prostate
Hedgehog
SMO
GDC-0449
Mutation
(PTCH/SMO)
18/33 (56%) Basal
Cell
EML4-ALK
PF-02341066
Translocation
20/31 (61%) Lung
BRAFV600E
PLX4032 (RG7204)
Mutation
19/27 (70%)
Melanoma
Fong et al NEJM, 2009; von Hoff et al NEJM 2009; Kwak et al ECCO/ESMO
2009: Chapman et al ECCO/ESMO 2009;
18
Biomarkers in Phase 2/3
• Focus on developing predictive markers
• Difficult to demonstrate that the absence of
predictive markers contributed to “failure” of
drug
◦ Known prior to phase III
 HER2,
◦ Positive phase III subsequently analyzed for subset and
marker was helpful
 Cetuximab, panitumumab and KRAS
 Erlotinib/gefitinib and EGFR FISH and mutations
Or not
 EGFR IHC in colon or lung carcinoma
◦ Negative phase III not further evaluated or under
evaluation
19
Phase 3 (or 2) Trial: Effects of Biomarker Assay
Initial
Selection
Strata
Agent
Marker +
Histology
Stage
Randomize
Target
Tested
Marker -
Control
Agent
Outcome
Phase 3:
Survival
(Phase 2:
ORR, TTE)
Control
• Trial is designed to assess treatment effects in Marker+ and Markergroups
• Marker assessment
◦ Assay failure increases number of patients screened
◦ False positives will dilute effect
◦ False negatives will increase the number of patients screened
20
Phase 3 (or 2) Trial –Effect of Assay False Positive and
Negatives
Marker+
Treatment+
Marker-
Treatment Marker x Treatment
Interaction with False Assay Results
Ideal Marker x Treatment
Interaction
Survival
Survival
M+/T+
M+/T+
M-/T+
M+/TM-/TM-/T+
Time
21
Suppose we have a new targeted therapy designed to
be effective in patients with Marker A.
What types of clinical trials should we design?
22
Biomarker Clinical Trial Designs
• Target Selection or Enrichment Designs
• Unselected or All-comers designs
◦ Marker by treatment interaction designs (biomarker
stratified design)
◦ Adaptive analysis designs
◦ Sequential testing strategy designs
◦ Biomarker-strategy designs
• Hybrid designs
Target Selection/Enrichment Designs
If we are sure that the therapy will not work in Markernegative patients
AND
We have an assay that can reliably assess the Marker
THEN
We might design and conduct clinical trials for Markerpositive patients or in subsets of patients with high
likelihood of being Marker-positive
IPASS-Schema
East Asian
Never smoker/light
former smoker
Pulmonary
Adenocarcinoma
No prior treatment
R
A
N
D
O
M
I
Z
E
Mok et al N Engl J Med 2009;361:947-57
Gefitinib
250 mg daily
Paclitaxel 200 mg/m2
Carboplatin AUC 5-6
1° Endpoint PFS
2° EGFR Biomarker 25
IPASS-Gefitinib or Carboplatin–Paclitaxel in Pulmonary
Adenocarcinoma.
Mok et al N Engl J Med 2009;361:947-57
26
Unselected “All Comers” Trial Designs
If we are not sure that the Marker will define groups of
patients that will benefit/not benefit from treatment
OR
There isn’t a validated assay that can reliably assess the
status of the Marker
THEN
We might design and conduct clinical trials in unselected
patients and try to identify predictive markers and robust
assays.
Retrospective and Prospective Analysis Designs
Retrospective Analyses Designs
• Hypothesis generation studies
◦
Retrospective analyses based on convenience samples
• Prospective/retrospective designs
Prospective Designs
• Marker by treatment interaction designs (biomarker stratified
design)
• Adaptive analysis designs
• Biomarker-strategy designs
• Sequential testing strategy designs
• Hybrid designs
Prospective/Retrospective Design
• Well-conducted randomized controlled trial
• Prospectively stated hypothesis, analysis techniques,
and patient population
• Predefined and standardized assay and scoring system
Prospective
• Upfront sample size and power calculation
• Samples collected during trial and available on a large
majority of patients to avoid selection bias
• Biomarker status is evaluated after the analysis of
clinical outcomes
Retrospective
• Results are confirmed by independent RCT(s)
29
Suppose we want to find out if using a
biomarker to select treatment is better?
30
Marker-based Strategy Design
If we think that one therapy will work in Marker-negative and
another therapy will work in the Marker-positive patients
AND
We have a validated assay that can reliably assess the Marker
status
THEN
We might design and conduct clinical trial to test whether using
the biomarker to select treatment for patients is better than
not using the marker to select treatment
Marker-based Strategy Design
Marker-Guided Randomized Design
Randomize To Use Of Marker Versus No Marker Evaluation
Control patients may receive standard or be randomized
All Patients
Randomize
Marker Determined
Treatment
M+
New Drug
M−
Control
New Drug
Randomize Treatment
OR
Standard Treatment
Control
Control
• Provides measure of patient willingness to follow marker-assigned therapy
• Marker guided treatment may be attractive to patients or clinicians
• Inefficient compared to completely randomized or randomized block design
32
Example: ERCC1: Customizing Cisplatin Based on
Quantitative Excision Repair Cross-Complementing 1
mRNA Expression
Cobo M et al. J Clin Oncol; 25:2747-2754 2007
•
•
•
444 chemotherapy-naïve patients with stage IIIB/IV NSCLC enrolled,
78 (17.6%) went off study before receiving chemotherapy, due insufficient tumor for
ERCC1 mRNA assessment.
346 patients assessable for response: Objective response was 39.3% in the control
arm and 50.7% in the genotypic arm (P = .02).
33
Predictive Markers Trials: Considerations
• Is the drug/treatment active?
• Do we have a marker/markers?
• What are the treatment effects within patient subsets?
◦ Are there enough patients to assess treatment effects in
Marker+ and/or Marker- groups?
• Does the trial design distinguish predictive and prognostic
effects?
• Is there a reliable assay to assess the biomarker?
• Samples requirements
34
Biomarker Translational Gaps
Laboratory
Single Centre
Trial
Multi-Centre
Trial
Clinical
Practice
Rapidity of
Science
Technology
Operations
Regulation
Standardization
Impact
•
•
•
Rapid generation of new science in laboratory
High content single institution trials can address biological questions
Impact requires translation to multi-centre trials and ultimately clinical
practice
35
Unprecedented Opportunity
• Rapid advances in understanding of cancer
biology
• Rapid advances in technology
• An increasing arsenal of active agents available
commercially or under clinical development
• Many opportunities for biomarker evaluation
36
8 Considerations for Biomarkers in Clinical Trials
• What is the question?
• Biomarker(s) – What we want to measure
• Assay – How we measure it
• Specimen – What we measure it in
• Study/Trial Design – Why, when, how we study it
• Study Execution – Can we get the study done
• Study Outcome – What it tells us
• Likely Impact – Whether we use it
37