harvard&catalyst adap.ve&clinical&trials& design&and ... · definition of an...
TRANSCRIPT
Harvard Catalyst Adap.ve Clinical Trials
Design and Implementa.on
Zoran Antonijevic
• Overview of Novel Adap.ve Designs • Implementa.on • DMC issues
Cytel 2013 Harvard Catalyst 31May2013 2
Presentation Contents
Cytel 2013 Harvard Catalyst 31May2013 3
Overview of Novel Adap.ve Designs
Definition of an Adaptive Design
• A study design is called "Adap.ve" if it allows modifica.on of an essen.al design feature (e.g., sample size, randomiza.on ra.o, number of treatment arms), based on accruing data from within that clinical trial – Should be carried out without compromising the integrity of the trial
– For confirmatory trials full details of adapta.ons need to be pre-‐specified
Cytel 2013 4 Harvard Catalyst 31May2013
• Also known as “Less-‐Well Understood” Adap.ve Designs (DraX FDA AD Guidance):
• Phase 2 – Adap.ve Dose Selec.on Studies
• Phase 3 – Unblinded Sample Size Re-‐Assessment – Confirmatory Trials with Dose Selec.on (Seamless Phase 2b/3)
– Confirmatory Trials with Sub-‐Popula.on selec.on (Adap.ve Enrichment Design)
Cytel 2013 Harvard Catalyst 31May2013 5
Novel Adaptive Designs
When applicable: entering Phase 3 there is s.ll residual uncertainty regarding the treatment effect
GSD vs. Unblined SSR: 1. Both address the same
problem 2. With GSD one commits to a
larger study ini.ally, can stop early if efficacy becer than assumed, or for fu.lity
3. With USSR one ini.ally assumes a more op.mis.c effect, increases the sample size at the interim as needed.
Unblinded Sample Size Re-estimation
2 Harvard Catalyst 31May2013
When applicable: entering Phase 3 there isn’t enough evidence that Biomarker subpopula.on would perform (substan.ally) becer than complementary (or full) popula.on
Randomiza.on stra.fied by the Biomarker status. At the interim the enrolment in the complementary sub may be stopped as guided by the data.
IA50%
IA70%
Stop for E!cacy
Stop for Futility
Continueas Planned
IncreaseSample Size
EnrichPopulation
Confirmatory Trial with Sub-population Selection (Adaptive Enrichment)
2 Harvard Catalyst 31May2013
When applicable: aXer comple.on of Phase2 there is s.ll residual uncertainty regarding the best dose Formerly known as “Seamless Phase 2b/3, and has caused some controversy. Now actually recommended in the guidance.
Preferably ini.ated as a follow-‐up to Phase 2b, although many sponsors s.ll use it as a replacement for (or skipping) Phase 2b.
2
Confirmatory Trial with Dose Selection
Harvard Catalyst 31May2013
Cytel 2013 Harvard Catalyst 31May2013 9
Implementa.on
• Required exper.se – Sta.s.cal methodology – Independent data monitoring commicee (IDMC) involving both clinicians and
sta.s.cians – Experienced opera.onal team to manage the logis.cal complexi.es and firewalls to
limit poten.al opera.onal biases • Adequate planning .me – Pre-‐study simula.ons – Early communica.on with the regulators for confirmatory trials (e.g. SPA, Parallel
Scien.fic Advice -‐ FDA/EMA) – Document prepara.on (DMC Charter and SAP submission) and opera.onal setup • Choosing the right technology – IVRS and/or IWRS: Drug logis.c management and randomiza.on – EDC: Real-‐.me access to high-‐quality data – Endpoint Adjudica.on: Centralized processing and independent review – Integrated Workflow Management System: Strict security levels (firewalls) to ensure
data and trial integrity in compliance with regulatory guidance Cytel 2013 Harvard Catalyst 31May2013 10
Key Considerations when Planning Adaptive Studies
• Key issues: – Who knows what and when? – Access control and verifica.on – Firewalls – Storage of DMC materials – Communica.on of Results and Recommenda.ons
Cytel 2013 11
Data monitoring of Adaptive Clinical Trials
Harvard Catalyst 31May2013
Cytel 2013 12
Improving the Process
ISC
DMC
Sponsor
EDC IRT CTMS
Harvard Catalyst 31May2013
Cytel 2013 Harvard Catalyst 31May2013 13
DMC Issues
• The FDA Adap.ve Design Guidance: “Because the DMC is unblinded to interim study results it can help implement the adapta2on decision according to the prospec2ve adapta2on algorithm, but it should not be in a posi.on to otherwise change the study design except for serious safety-‐related concerns that are the usual responsibility of a DMC”
Cytel 2013 14
DMC Role
Harvard Catalyst 31May2013
• Similari.es with GSD: -‐ The adapta2on algorithm can be pre-‐specified in a rela2vely straigh>orward manner
-‐ Both based on the unblinded results of the treatment effect -‐ The pre-‐specified sta.s.cal algorithm can be compromised by safety findings for both of these designs
• Differences: – GSD; concerns re overflow – SSR; design carefully not to reveal treatment effect
Cytel 2013 15
Unblinded Sample Size Reassessment
Harvard Catalyst 31May2013
• Decisions involve parameters beyond just ethical considera.ons. • Include decisions that tradi.onally have been sponsor
responsibili.es. • May have major commercial implica.ons, for example dose
selec.on or sub-‐popula.on selec.on. • Complex algorithms • DMC may desire to consult with the steering commicee that
includes a sponsor representa.ve not involved in trial management.
Cytel 2013 16
Complex Adaptive Designs at Confirmatory Stage
Harvard Catalyst 31May2013
• Not intended to demonstrate confirmatory evidence; less rigor in restric.ng knowledge of interim results
• Lesser regulatory implica.ons, though avoiding opera.onal bias should be recommended in any trial
• May require frequent updates based on interim data • Adapta.ons are oXen driven by a pre-‐specified algorithm
automated through an integrated response system • Therefore, it may be preferable that recommenda.ons are
endorsed by an Internal Review Commicee
Cytel 2013 17
Adaptive Designs in Exploratory Studies
Harvard Catalyst 31May2013
• US Food and Drug Administra.on. Guidance for industry: adap.ve design clinical trials for drugs and biologics (draX guidance). February 2010. Available online at hcp://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInforma.on/Guidances/UCM201790.pdf
Cytel 2013 Harvard Catalyst 31May2013 18
References
Cytel 2013 Harvard Catalyst 31May2013 19
Back-‐up Slides
Value of Design • Easily under-‐appreciated, trial designs have a major impact
on the value of a pharmaceu.cal product
Adap.ve Phase 2 Designs • The FDA AD Guidance recommends use of AD in Phase 2 • AD is the most efficient approach for this stage of
development
Development Program Op.miza.on • Phase 2 most cri.cal for op2mal drug development
Development Scenario Simula.on • Comparison of rela.ve efficiency of various development
op2ons can be accomplished only with simula2ons
Harvard Catalyst 31May2013 20
Phase 2b Dose Finding Studies
Cytel 2013
Scenarios – (7) different sta.s.cal approaches – (2) Phase 2 strategies – (3) Phase 3 strategies – For a total of 42 scenarios. Same assump.ons for safety and efficacy used for all scenarios
– Cost and development .me calculated using real data from exis.ng databases
Results – The expected NPV for different scenarios ranged from $0.5B to $1.2B
– Adap.ve Designs far outperformed tradi.onal approaches
Case Study
Harvard Catalyst 31May2013 21 Cytel 2013
Expected NPV
22 Average NPV (millions)
ANOVA
Dopt
GADA
MCPMod
MTT
BMA
LOCFIT
600 800 1000
N = 150
ANOVA
Dopt
GADA
MCPMod
MTT
BMA
LOCFIT
N = 250
1 dose 2 doses, fast 2 doses, normal
Harvard Catalyst 31May2013 Cytel 2013
Method