Adaptive designs - an overview and some examples · Motivation Adaptive Designs Examples How and...
Transcript of Adaptive designs - an overview and some examples · Motivation Adaptive Designs Examples How and...
Adaptive designs - an overview and someexamples
Thomas Jaki Philip PallmannLancaster University
MotivationAdaptive Designs
Examples
Attrition rates for new developments (Arrowsmith 2011a, 2011b)
phase II: >80%
phase III & submission: ∼50%
Reasons for failure (Arrowsmith & Miller 2013)
Phase II (2011–2012)
Efficacy Safety Other
Phase III & submission (2011–2012)
Efficacy Safety Other
c© MPS Research Unit Adaptive trials 2
MotivationAdaptive Designs
Examples
Attrition rates for new developments (Arrowsmith 2011a, 2011b)
phase II: >80%
phase III & submission: ∼50%
Reasons for failure (Arrowsmith & Miller 2013)
Phase II (2011–2012)
Efficacy Safety Other
Phase III & submission (2011–2012)
Efficacy Safety Other
c© MPS Research Unit Adaptive trials 3
MotivationAdaptive Designs
Examples
Attrition rates for new developments (Arrowsmith 2011a, 2011b)
phase II: >80%
phase III & submission: ∼50%
Reasons for failure (Arrowsmith & Miller 2013)
Phase II (2011–2012)
Efficacy Safety Other
Phase III & submission (2011–2012)
Efficacy Safety Other
c© MPS Research Unit Adaptive trials 4
MotivationAdaptive Designs
Examples
Attrition rates for new developments (Arrowsmith 2011a, 2011b)
phase II: >80%
phase III & submission: ∼50%
Reasons for failure (Arrowsmith & Miller 2013)
Phase II (2011–2012)
Efficacy Safety Other
Phase III & submission (2011–2012)
Efficacy Safety Other
c© MPS Research Unit Adaptive trials 5
MotivationAdaptive Designs
Examples
Attrition rates for new developments (Arrowsmith 2011a, 2011b)
phase II: >80%
phase III & submission: ∼50%
Reasons for failure (Arrowsmith & Miller 2013)
Phase II (2011–2012)
Efficacy Safety Other
Phase III & submission (2011–2012)
Efficacy Safety Other
c© MPS Research Unit Adaptive trials 6
MotivationAdaptive Designs
Examples
Likely causes for failure:
taking forward futile treatments
studying the wrong patient population
poor precision (optimal dose, maximum tolerated dose, safety)
c© MPS Research Unit Adaptive trials 7
MotivationAdaptive Designs
Examples
Likely causes for failure:
taking forward futile treatments
studying the wrong patient population
poor precision (optimal dose, maximum tolerated dose, safety)
c© MPS Research Unit Adaptive trials 8
MotivationAdaptive Designs
Examples
Likely causes for failure:
taking forward futile treatments
studying the wrong patient population
poor precision (optimal dose, maximum tolerated dose, safety)
c© MPS Research Unit Adaptive trials 9
MotivationAdaptive Designs
Examples
Likely causes for failure:
taking forward futile treatments
studying the wrong patient population
poor precision (optimal dose, maximum tolerated dose, safety)
c© MPS Research Unit Adaptive trials 10
MotivationAdaptive Designs
Examples
Can we do better?
avoid going straight into large and expensive phase III
take more care during phases I and II
consider adaptive and Bayesian designs
c© MPS Research Unit Adaptive trials 11
MotivationAdaptive Designs
Examples
Can we do better?
avoid going straight into large and expensive phase III
take more care during phases I and II
consider adaptive and Bayesian designs
c© MPS Research Unit Adaptive trials 12
MotivationAdaptive Designs
Examples
Can we do better?
avoid going straight into large and expensive phase III
take more care during phases I and II
consider adaptive and Bayesian designs
c© MPS Research Unit Adaptive trials 13
MotivationAdaptive Designs
Examples
Can we do better?
avoid going straight into large and expensive phase III
take more care during phases I and II
consider adaptive and Bayesian designs
c© MPS Research Unit Adaptive trials 14
MotivationAdaptive Designs
Examples
2. Adaptive Designs
c© MPS Research Unit Adaptive trials 15
MotivationAdaptive Designs
Examples
Idea
Modify an ongoing trial
by design or ad hoc
based on reviewing accrued data at interim
to enhance flexibility
without undermining the study’s integrity and validity.(Chow et al. 2005)
c© MPS Research Unit Adaptive trials 16
MotivationAdaptive Designs
Examples
Idea
Modify an ongoing trial
by design or ad hoc
based on reviewing accrued data at interim
to enhance flexibility
without undermining the study’s integrity and validity.(Chow et al. 2005)
c© MPS Research Unit Adaptive trials 17
MotivationAdaptive Designs
Examples
Idea
Modify an ongoing trial
by design or ad hoc
based on reviewing accrued data at interim
to enhance flexibility
without undermining the study’s integrity and validity.(Chow et al. 2005)
c© MPS Research Unit Adaptive trials 18
MotivationAdaptive Designs
Examples
Idea
Modify an ongoing trial
by design or ad hoc
based on reviewing accrued data at interim
to enhance flexibility
without undermining the study’s integrity and validity.(Chow et al. 2005)
c© MPS Research Unit Adaptive trials 19
MotivationAdaptive Designs
Examples
Idea
Modify an ongoing trial
by design or ad hoc
based on reviewing accrued data at interim
to enhance flexibility
without undermining the study’s integrity and validity.(Chow et al. 2005)
c© MPS Research Unit Adaptive trials 20
MotivationAdaptive Designs
Examples
Pros and cons
+ highly flexible
+ very efficient
+ reflects medical practice
+ shorter trial and/or moreaccurate estimates
+ ethical
– highly flexible
– inefficient
– time-consuming to design
– simple estimates may be biased
– interim analyses may requireunblinding
c© MPS Research Unit Adaptive trials 21
MotivationAdaptive Designs
Examples
Fixed sample design
Plan studyFix n
Start Final analysis
total sample size known in advance
no adjustment possible
c© MPS Research Unit Adaptive trials 22
MotivationAdaptive Designs
Examples
Fixed sample design
Plan studyFix n
Start Final analysis
total sample size known in advance
no adjustment possible
c© MPS Research Unit Adaptive trials 23
MotivationAdaptive Designs
Examples
Fixed sample design
Plan studyFix n
Start Final analysis
total sample size known in advance
no adjustment possible
c© MPS Research Unit Adaptive trials 24
MotivationAdaptive Designs
Examples
(Group-)sequential design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
larger maximum sample size
lower expected sample size
c© MPS Research Unit Adaptive trials 25
MotivationAdaptive Designs
Examples
(Group-)sequential design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
larger maximum sample size
lower expected sample size
c© MPS Research Unit Adaptive trials 26
MotivationAdaptive Designs
Examples
(Group-)sequential design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
larger maximum sample size
lower expected sample size
c© MPS Research Unit Adaptive trials 27
MotivationAdaptive Designs
Examples
(Group-)sequential design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
larger maximum sample size
lower expected sample size
c© MPS Research Unit Adaptive trials 28
MotivationAdaptive Designs
Examples
(Group-)sequential design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
larger maximum sample size
lower expected sample size
c© MPS Research Unit Adaptive trials 29
MotivationAdaptive Designs
Examples
(Group-)sequential design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
larger maximum sample size
lower expected sample size
c© MPS Research Unit Adaptive trials 30
MotivationAdaptive Designs
Examples
Adaptive design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
and many other options . . .
larger maximum sample size
lower expected sample size
possibly more relevant answer
c© MPS Research Unit Adaptive trials 31
MotivationAdaptive Designs
Examples
Adaptive design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
and many other options . . .
larger maximum sample size
lower expected sample size
possibly more relevant answer
c© MPS Research Unit Adaptive trials 32
MotivationAdaptive Designs
Examples
Adaptive design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
and many other options . . .
larger maximum sample size
lower expected sample size
possibly more relevant answer
c© MPS Research Unit Adaptive trials 33
MotivationAdaptive Designs
Examples
Adaptive design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
and many other options . . .
larger maximum sample size
lower expected sample size
possibly more relevant answer
c© MPS Research Unit Adaptive trials 34
MotivationAdaptive Designs
Examples
Adaptive design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
and many other options . . .
larger maximum sample size
lower expected sample size
possibly more relevant answer
c© MPS Research Unit Adaptive trials 35
MotivationAdaptive Designs
Examples
Adaptive design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
and many other options . . .
larger maximum sample size
lower expected sample size
possibly more relevant answer
c© MPS Research Unit Adaptive trials 36
MotivationAdaptive Designs
Examples
Adaptive design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
and many other options . . .
larger maximum sample size
lower expected sample size
possibly more relevant answer
c© MPS Research Unit Adaptive trials 37
MotivationAdaptive Designs
Examples
Adaptive design
Plan studyFix design
Start Interim analyses Final analysis
At each interim:
decide whether or not to stop
and many other options . . .
larger maximum sample size
lower expected sample size
possibly more relevant answer
c© MPS Research Unit Adaptive trials 38
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 39
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 40
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 41
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 42
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 43
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 44
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 45
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 46
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 47
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 48
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 49
MotivationAdaptive Designs
Examples
How and what can we adapt?
Prospectively (“by design")
adaptive randomization
early stopping (safety, futility, efficacy)
drop-the-loser(s), pick-the-winner
sample size re-estimation (to achieve desired power)
Concurrently (“ad hoc")
modify inclusion/exclusion criteria
change primary endpoint
modify doses
extend treatment duration
change trial objective (e.g., non-inferiority→ superiority)
c© MPS Research Unit Adaptive trials 50
MotivationAdaptive Designs
Examples
A single-stage design
ABANDON
PROCEED
n
Z
c© MPS Research Unit Adaptive trials 51
MotivationAdaptive Designs
Examples
Simon’s two-stage design
ABANDON
PROCEED
CONTINUE
n1 n
Z
(Simon 1989)
c© MPS Research Unit Adaptive trials 52
MotivationAdaptive Designs
Examples
A multi-stage design
ABANDON
PROCEED
CONTINUE
n1 n2 n3 n
Z
(Whitehead 1997)
c© MPS Research Unit Adaptive trials 53
MotivationAdaptive Designs
Examples
A multi-stage design
n1
Z
CONTINUEx
(Whitehead 1997)
c© MPS Research Unit Adaptive trials 54
MotivationAdaptive Designs
Examples
A multi-stage design
n1 n2
Z
CONTINUEx x
(Whitehead 1997)
c© MPS Research Unit Adaptive trials 55
MotivationAdaptive Designs
Examples
A multi-stage design
n1 n2 n3
Z
PROCEEDx x
x
(Whitehead 1997)
c© MPS Research Unit Adaptive trials 56
MotivationAdaptive Designs
Examples
Example 1: Sample size re-estimation
Problem:
sample size formulae depend on nuisance parameters e.g., σ2
Idea:
estimate the nuisance parameter(s) at interim and recalculate sample size
Methods:
for binary, Gaussian or survival outcomes
exact (with unblinding) or approximate (without unblinding) (Gould 1995)
negligible effect on type I error
c© MPS Research Unit Adaptive trials 57
MotivationAdaptive Designs
Examples
Example 1: Sample size re-estimation
Problem:
sample size formulae depend on nuisance parameters e.g., σ2
Idea:
estimate the nuisance parameter(s) at interim and recalculate sample size
Methods:
for binary, Gaussian or survival outcomes
exact (with unblinding) or approximate (without unblinding) (Gould 1995)
negligible effect on type I error
c© MPS Research Unit Adaptive trials 58
MotivationAdaptive Designs
Examples
Example 1: Sample size re-estimation
Problem:
sample size formulae depend on nuisance parameters e.g., σ2
Idea:
estimate the nuisance parameter(s) at interim and recalculate sample size
Methods:
for binary, Gaussian or survival outcomes
exact (with unblinding) or approximate (without unblinding) (Gould 1995)
negligible effect on type I error
c© MPS Research Unit Adaptive trials 59
MotivationAdaptive Designs
Examples
Example 1: Sample size re-estimation
Problem:
sample size formulae depend on nuisance parameters e.g., σ2
Idea:
estimate the nuisance parameter(s) at interim and recalculate sample size
Methods:
for binary, Gaussian or survival outcomes
exact (with unblinding) or approximate (without unblinding) (Gould 1995)
negligible effect on type I error
c© MPS Research Unit Adaptive trials 60
MotivationAdaptive Designs
Examples
Example 1: Sample size re-estimation
Problem:
sample size formulae depend on nuisance parameters e.g., σ2
Idea:
estimate the nuisance parameter(s) at interim and recalculate sample size
Methods:
for binary, Gaussian or survival outcomes
exact (with unblinding) or approximate (without unblinding) (Gould 1995)
negligible effect on type I error
c© MPS Research Unit Adaptive trials 61
MotivationAdaptive Designs
Examples
Example 1: Sample size re-estimation
Problem:
sample size formulae depend on nuisance parameters e.g., σ2
Idea:
estimate the nuisance parameter(s) at interim and recalculate sample size
Methods:
for binary, Gaussian or survival outcomes
exact (with unblinding) or approximate (without unblinding) (Gould 1995)
negligible effect on type I error
c© MPS Research Unit Adaptive trials 62
MotivationAdaptive Designs
Examples
Sample size re-estimationExample: DEVELOP-UK trial
transplantation of reconditioned vs. standard donor lungs
ex vivo lung perfusion (EVLP)
phase III, multi-centre, unblinded, non-randomised,non-inferiority observational study
primary endpoint: 12 months survival
uncertainty in design parameters (only 50 transplantsworldwide)
c© MPS Research Unit Adaptive trials 63
MotivationAdaptive Designs
Examples
Example 1: Sample size re-estimationDEVELOP-UK trial
408 patients randomised to EVLP and standard
3:1 in favour of standard to ensure all available lungs areused
interim analyses after 1/3 and 2/3 of total sample sizefirst: early stoppingsecond: early stopping, sample size re-assessment
significance levels: 0.005 (first), 0.014 (second), 0.045 (final)
c© MPS Research Unit Adaptive trials 64
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Dilemma:
low doses inefficient←→ high doses toxic
Goal:
find a maximum tolerated dose (MTD)where a prespecified proportion p0 of patients
experience dose-limiting toxicities (DLTs)
c© MPS Research Unit Adaptive trials 65
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Dilemma:
low doses inefficient←→ high doses toxic
Goal:
find a maximum tolerated dose (MTD)where a prespecified proportion p0 of patients
experience dose-limiting toxicities (DLTs)
c© MPS Research Unit Adaptive trials 66
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Designs:
3 + 3 design (Carter 1973)
. . . and variations (A + B, rolling 6, . . . )
other 1- or 2-stage up-and-down designs
accelerated titration design (Simon et al. 1999)
continual reassessment method (O’Quigley et al. 1990)
c© MPS Research Unit Adaptive trials 67
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Designs:
3 + 3 design (Carter 1973)
. . . and variations (A + B, rolling 6, . . . )
other 1- or 2-stage up-and-down designs
accelerated titration design (Simon et al. 1999)
continual reassessment method (O’Quigley et al. 1990)
c© MPS Research Unit Adaptive trials 68
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Designs:
3 + 3 design (Carter 1973)
. . . and variations (A + B, rolling 6, . . . )
accelerated titration design (Simon et al. 1999)
rule-based
continual reassessment method (O’Quigley et al. 1990) model-based
c© MPS Research Unit Adaptive trials 69
MotivationAdaptive Designs
Examples
Dose-finding / Phase I
Do NOT use rule based designs
c© MPS Research Unit Adaptive trials 70
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Continual reassessment method (CRM):
Bayesian adaptive design
model-based
synthesizes prior belief about MTD and accumulating data
uses information from all previous patients (“memory")
c© MPS Research Unit Adaptive trials 71
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Continual reassessment method (CRM):
Bayesian adaptive design
model-based
synthesizes prior belief about MTD and accumulating data
uses information from all previous patients (“memory")
c© MPS Research Unit Adaptive trials 72
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Continual reassessment method (CRM):
Bayesian adaptive design
model-based
synthesizes prior belief about MTD and accumulating data
uses information from all previous patients (“memory")
c© MPS Research Unit Adaptive trials 73
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Continual reassessment method (CRM):
Bayesian adaptive design
model-based
synthesizes prior belief about MTD and accumulating data
uses information from all previous patients (“memory")
c© MPS Research Unit Adaptive trials 74
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Continual reassessment method (CRM):
Bayesian adaptive design
model-based
synthesizes prior belief about MTD and accumulating data
uses information from all previous patients (“memory")
c© MPS Research Unit Adaptive trials 75
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
A CRM design:
Dose
p
p0
D1 D2 D3 D4
based on prior belief
p: proportion of DLTs
c© MPS Research Unit Adaptive trials 76
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
A CRM design:
Dose
p
p0
D1 D2 D3 D4
based on prior belief
p: proportion of DLTs
c© MPS Research Unit Adaptive trials 77
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
A CRM design:
Dose
p
p0
D1 D2 D3 D4
based on prior belief
p: proportion of DLTs
c© MPS Research Unit Adaptive trials 78
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
A CRM design:
Dose
p
p0
D1 D2 D3 D4
based on prior belief
p: proportion of DLTs
c© MPS Research Unit Adaptive trials 79
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
A CRM design:
Dose
p
p0
D1 D2 D3 D4
based on prior belief
based on prior belief & data
p: proportion of DLTs
c© MPS Research Unit Adaptive trials 80
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
A CRM design:
Dose
p
p0
D1 D2 D3 D4
based on prior belief
based on prior belief & data
p: proportion of DLTs
c© MPS Research Unit Adaptive trials 81
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
A CRM design:
Dose
p
p0
D1 D2 D3 D4
based on prior belief
based on prior belief & data
p: proportion of DLTs
c© MPS Research Unit Adaptive trials 82
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
CRM vs. 3 + 3:
more likely to identify the correct dose
more patients receive doses near MTD
fewer overdoses
clearly defined MTD
MTD estimated with a measure of precision
efficient use of all data
flexible choice of toxicity levels
easily extendable
c© MPS Research Unit Adaptive trials 83
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Seamless phase I/II designs:
phase I study “with a phase II flavour" (O’Quigley et al. 2001)
model toxicity and efficacy together
identify the most successful dose (MSD) rather than MTD
efficient (patient population of interest in phase I)
various methods using different models and stopping rules:
efficacy-toxicity trade-off (Thall & Russell 1998, Thall & Cook 2004)
CRM-like (O’Quigley et al. 2001, Zohar & O’Quigley 2006)
decision procedure for continuous response (Zhou et al. 2006)
c© MPS Research Unit Adaptive trials 84
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Seamless phase I/II designs:
phase I study “with a phase II flavour" (O’Quigley et al. 2001)
model toxicity and efficacy together
identify the most successful dose (MSD) rather than MTD
efficient (patient population of interest in phase I)
various methods using different models and stopping rules:
efficacy-toxicity trade-off (Thall & Russell 1998, Thall & Cook 2004)
CRM-like (O’Quigley et al. 2001, Zohar & O’Quigley 2006)
decision procedure for continuous response (Zhou et al. 2006)
c© MPS Research Unit Adaptive trials 85
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Seamless phase I/II designs:
phase I study “with a phase II flavour" (O’Quigley et al. 2001)
model toxicity and efficacy together
identify the most successful dose (MSD) rather than MTD
efficient (patient population of interest in phase I)
various methods using different models and stopping rules:
efficacy-toxicity trade-off (Thall & Russell 1998, Thall & Cook 2004)
CRM-like (O’Quigley et al. 2001, Zohar & O’Quigley 2006)
decision procedure for continuous response (Zhou et al. 2006)
c© MPS Research Unit Adaptive trials 86
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Seamless phase I/II designs:
phase I study “with a phase II flavour" (O’Quigley et al. 2001)
model toxicity and efficacy together
identify the most successful dose (MSD) rather than MTD
efficient (patient population of interest in phase I)
various methods using different models and stopping rules:
efficacy-toxicity trade-off (Thall & Russell 1998, Thall & Cook 2004)
CRM-like (O’Quigley et al. 2001, Zohar & O’Quigley 2006)
decision procedure for continuous response (Zhou et al. 2006)
c© MPS Research Unit Adaptive trials 87
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Seamless phase I/II designs:
phase I study “with a phase II flavour" (O’Quigley et al. 2001)
model toxicity and efficacy together
identify the most successful dose (MSD) rather than MTD
efficient (patient population of interest in phase I)
various methods using different models and stopping rules:
efficacy-toxicity trade-off (Thall & Russell 1998, Thall & Cook 2004)
CRM-like (O’Quigley et al. 2001, Zohar & O’Quigley 2006)
decision procedure for continuous response (Zhou et al. 2006)
c© MPS Research Unit Adaptive trials 88
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Seamless phase I/II designs:
phase I study “with a phase II flavour" (O’Quigley et al. 2001)
model toxicity and efficacy together
identify the most successful dose (MSD) rather than MTD
efficient (patient population of interest in phase I)
various methods using different models and stopping rules:efficacy-toxicity trade-off (Thall & Russell 1998, Thall & Cook 2004)
CRM-like (O’Quigley et al. 2001, Zohar & O’Quigley 2006)
decision procedure for continuous response (Zhou et al. 2006)
c© MPS Research Unit Adaptive trials 89
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Seamless phase I/II designs:
phase I study “with a phase II flavour" (O’Quigley et al. 2001)
model toxicity and efficacy together
identify the most successful dose (MSD) rather than MTD
efficient (patient population of interest in phase I)
various methods using different models and stopping rules:efficacy-toxicity trade-off (Thall & Russell 1998, Thall & Cook 2004)
CRM-like (O’Quigley et al. 2001, Zohar & O’Quigley 2006)
decision procedure for continuous response (Zhou et al. 2006)
c© MPS Research Unit Adaptive trials 90
MotivationAdaptive Designs
Examples
Example 2: Phase I cancer trials
Seamless phase I/II designs:
phase I study “with a phase II flavour" (O’Quigley et al. 2001)
model toxicity and efficacy together
identify the most successful dose (MSD) rather than MTD
efficient (patient population of interest in phase I)
various methods using different models and stopping rules:efficacy-toxicity trade-off (Thall & Russell 1998, Thall & Cook 2004)
CRM-like (O’Quigley et al. 2001, Zohar & O’Quigley 2006)
decision procedure for continuous response (Zhou et al. 2006)
c© MPS Research Unit Adaptive trials 91
MotivationAdaptive Designs
Examples
Example 3: Multi-arm multi-stage trials (MAMS)TAILoR trial (Pushpakom et al. 2015)
Phase II study to evaluate treatment for side effect of HIVtri-regimen treatment (TAILoR)
Superiority trial
Several possible doses
Normal distributed endpoint
c© MPS Research Unit Adaptive trials 92
MotivationAdaptive Designs
Examples
Example 3: Multi-arm multi-stage trials (MAMS)Design (Magirr et al, 2012)
3 active arms compared to control
Equal randomization between actives and control
one interim analysis
Stop for superiortiyStop if no active arm appears promisingDrop any active arms that are not sufficiently promising
In the trial we underestimated drop-out and could adjust atinterim
c© MPS Research Unit Adaptive trials 93
MotivationAdaptive Designs
Examples
General considerations
New methods either get results faster or are more accurate
Tend to take more time to develop upfront
How can this cost be covered?
Are the ideas sensible to implement?
Is an interim analysis feasible?Can Bayesian methods be updated online?
c© MPS Research Unit Adaptive trials 94