Traditional vs novel trial designing . Giuseppe M.C. Rosano, MD, PhD
Jan 14, 2016
Traditional vs novel trial designing
Giuseppe MC Rosano MD PhD
Loscalzo J (2012) Circulation 125638-645
Why should we do it faster
From 2000 to 20010 the top 10 pharma companies started 1113090more than 1700 new projectsyear
Why should we do it faster
bull Time
bull Costs
bull Efficiency
bull Innovation
25 Years life-cycle of a drug
New chemical entity approval and cost of RampD
13B$Drug
From bench to bedside
Reasons for termination of clinical development program on NCEs
C K Atterwill M G Wing Molecular and Cellular Approaches to Lead Optimization in Pharmaceutical Development
Definitions of stages of clinical trials
2B studies in selected indications
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Loscalzo J (2012) Circulation 125638-645
Why should we do it faster
From 2000 to 20010 the top 10 pharma companies started 1113090more than 1700 new projectsyear
Why should we do it faster
bull Time
bull Costs
bull Efficiency
bull Innovation
25 Years life-cycle of a drug
New chemical entity approval and cost of RampD
13B$Drug
From bench to bedside
Reasons for termination of clinical development program on NCEs
C K Atterwill M G Wing Molecular and Cellular Approaches to Lead Optimization in Pharmaceutical Development
Definitions of stages of clinical trials
2B studies in selected indications
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Why should we do it faster
bull Time
bull Costs
bull Efficiency
bull Innovation
25 Years life-cycle of a drug
New chemical entity approval and cost of RampD
13B$Drug
From bench to bedside
Reasons for termination of clinical development program on NCEs
C K Atterwill M G Wing Molecular and Cellular Approaches to Lead Optimization in Pharmaceutical Development
Definitions of stages of clinical trials
2B studies in selected indications
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
25 Years life-cycle of a drug
New chemical entity approval and cost of RampD
13B$Drug
From bench to bedside
Reasons for termination of clinical development program on NCEs
C K Atterwill M G Wing Molecular and Cellular Approaches to Lead Optimization in Pharmaceutical Development
Definitions of stages of clinical trials
2B studies in selected indications
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
New chemical entity approval and cost of RampD
13B$Drug
From bench to bedside
Reasons for termination of clinical development program on NCEs
C K Atterwill M G Wing Molecular and Cellular Approaches to Lead Optimization in Pharmaceutical Development
Definitions of stages of clinical trials
2B studies in selected indications
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
From bench to bedside
Reasons for termination of clinical development program on NCEs
C K Atterwill M G Wing Molecular and Cellular Approaches to Lead Optimization in Pharmaceutical Development
Definitions of stages of clinical trials
2B studies in selected indications
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Reasons for termination of clinical development program on NCEs
C K Atterwill M G Wing Molecular and Cellular Approaches to Lead Optimization in Pharmaceutical Development
Definitions of stages of clinical trials
2B studies in selected indications
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Definitions of stages of clinical trials
2B studies in selected indications
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
TARGET POPULATION
RANDOM
OUTCOMEOUTCOME
TREATMENT BTREATMENT A
STATISTICAL COMPARISON
Traditional controlled clinical trial
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Traditional trial designs
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Traditional trial designs
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Traditional trial designs
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Scheme for a prototypical n-of-1 trial
modified from Zucker et al 2006
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Animal models and human PKPDAnimal models and human PKPD
1 More than 25 of drug entering CTs fail because of ldquopoor pharmacologyrdquo
2 Animal data are useless to indicate the effective dose in humans (1100th of the therapeutic dose in animals)
bull Animal models are only required to test safety Efficacy should be either tested in big animals or should be directly tested in humans
3 Animal studies
bull Kill promising drugs (may work at therapeutic concentrations)
bull Promote drugs that are ineffective
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Can we skip earlier phases of clinical drug development
bull Aloperidol
bull LCZ696
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Adaptive trial designs bull A decision-oriented sequential learning process
bull At any stage the data may be analysed and next stages redesigned taking into account all available data from the trial or even from outside based on predefined rules
1 Allocation Rule how new patients will be assigned to available treatments
2 Sampling Rule how many subjects will be sampled at the next stage
3 Stopping Rule when to stop the trial
4 Decision Rule allows for change in hypothesis primary end-point statistical method or patient population design
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Adaptive trial designs
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Drop-Loser Adaptive Study Design
Learning phaseLearning phase Confirmatory phaseConfirmatory phase
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Adaptive Licensing the way forward
(a) Current scenario(b) AL scenario
Basic principleMake innovative drug access easier through an early approval and acknowledgement of uncertainty of data
Eichler et al Clin Pharmacol Ther 2012 Mar
hellipbut not of the risk
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
bull DurationDuration
bull CostsCosts
bull It is not efficientIt is not efficient
bull It is unnecessarily complicatedIt is unnecessarily complicated
bull Too many people InvolvedToo many people Involved
How to streamline clinical trialsHow to streamline clinical trials
Study design
Sample size
PI
CROs
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword
Patient recruitmentretention
1 Pre-trial registries
2 Consortiums
1 Use of social media (twitterfacebook)
2 New software
3 Patients organizations ndash double-edge sword