Transcript Slide 1

Oncology Section
RAW DRAFT
Andrew Buckler (section leader), Libero Marzella,
James Mountz, James Mulshine, Sonia Pearson-White,
Lawrence Schwartz, Barry Siegel, Annick Van den
Abbeele, Jeffrey Yap, Ricardo Avila, John Boone,
Michael Graham, David Gustafson, Edward Jackson,
Gary Kelloff, Paul Kinahan
Vision
Cancer will either be curable, or manageable as a chronic
sub-acute disease, and imaging will be central to its
management through the use of validated IB.
Vision 2025
 Socio-economic benefit
 Paradigm
 Clinical focus/relevance
 Performance
 Roadmap
Socio-economic benefit
(cheaper/better/faster)
1.
Less expensive/time consuming development:

Lower delta t and/or N for same statistical power
2. Manage cancer as chronic (personalized, sub-acute)
disease
 Clinical care (opportunity for strategic intervention,
reduction in errors, side effects, survival)
3. Pt is treated correctly for their disease.
Paradigm
 On the day you are born, health management begins (prediction)
 Prevent, and treat early disease (not late)
 Prevention strategies well developed, i.e., avoidable, reversible,
manageable
 Every patient that comes in the door will give a piece of themselves
from which ‘omic data are derived
 Deal effectively with heterogeneity
 Early stratification of patients at risk

Image is among a wide array of complementary assays that are well integrated
 Therapy is non-toxic and completely local
 Not just systemic treatments, include targeted minimally invasive
surgical, ablative, brachytherapy, etc.
 Change patient experience, error prone, expensive… more efficient
scans and diagnosis (pay for performance)
 Bioinformatics infrastructure robust and clinical data ubiquitous
Clinical focus/relevance
 Imaging for least intrusive window into pre-symptomatic phase of
disease
 Effective imaging signatures for hallmarks of pre-symptomatic disease
(both resolution and sensitivity)
 “virtual biopsy”
 Complementary benefit of other biomarkers need effective data fusion
approaches (incl. decision support)
 Genotyping->prevention (risk markers)
 Correlations between modality
 Imaging used both to stratify and to make adaptive therapy
personalized to pt
 Tx is highly localized, image guided, no side effects
Ask better questions: move full community beyond just asking whether we
can detect a tumor
Performance (the science of measurement
as opposed to picture taking)
 Standardization and iterative improvement:
 Variance reduced:
Human (e.g., protocols, compliance, structured reporting with quantitative endpoints)
 Sigma reduced through technology developments
 Calibration (incl. public data to calibrate against) (phantoms)
 CAD (as a way to reduce variance)
 Statistical framework for qualification established

 Multi-disciplinary approach to validation/qualification of imaging methods



Device vendors (both HW and SW) market product certified against qualified Profiles
 Qualification data shared across vendors 510ks and biopharma NDAs (cross registration)
(conformance checks established)
Results easily accessible (includes unsuccessful agents, not just positive)
Cancer centers should have conformance checking process, too
 Practice/workflow accounts for QI
 Methods worked out that make it easy to do (CAD)
 All imaging results embedded in EMR
Clinical imaging = Quantitative imaging
Roadmap: Year 1
 Start incorporating QI in clinical trial approval process ,
first candidate markers used in trial designs (contributed
by UPICT and QIBA)
 Build infrastructure in comunications:
 Approach to data sharing (needs to be linked to national agency initiatives)
(build linkage to caBIG, etc.) (needs to be able for regulatory agencies to
access and use)
 Establish lines of communication (incl. celebrate models of success) (incl.
strategy re: publishing)
 Engage ASCO, ASTRO,
 Engage top-management of device companies and biopharma
 Include QI in RFAs from NIH
 Define regulatory approval pathway (target: February 2010
meeting)
 Continue IRAT (and similar) through extramural funding
Roadmap: Year 3
 Establish the qualification model (that is, Integrating the
Imaging Biomarker Enterprise “IIBE”) (what QIBA is
actually working on)
 First marker through qualification process (a NDA/PMA type process) (e.g.,
FLT, vCT) (the “predicate”)
 Dynamic imaging with PK/PD analysis in place (e.g.,
hypoxia, angiogenesis, apoptosis, metabolism) (from
anatomic/structural to functional/dynamic) begun
 Engage CMMS
 Variance reduction efforts active and understood
 Incl. standardization of image acquisition, and processing
 Process of defining decision trees based on QI established and in use
(structured reporting)
Roadmap: Years 5-10
5
 First commercial product shipped with certified marker via 510(k)





(using the cross-reference)
Make quantitative imaging part of every appropriate clinical trial
Get payers to reimburse QI
QIN up and running
Structured reporting commonly used
Clinicians credentialed and “pulling”

Also it is in device tenders with buyers making decisions on it
 10
 Dynamic imaging with PK/PD analysis in common use (e.g.,
hypoxia, angiogenesis, apoptosis, metabolism) (from
anatomic/structural to functional/dynamic)
 National informatics infrastructure in place to support full
quantitative biomarker with meta-data
Cancer 2025
Mature
Emerg
ent
•Preclinical
•Clinical trials
•Screening
•Clinical care
•PK/PD
routinely
assessed with
imaging
commonly
used
•Models for
translational
medicine
(crossspecies)
more
characterize
d
•Metabolic agents and
receptor-based agents
are qualified for use
(e.g., VEGF)
•Proliferation,
apoptosis, hypoxia, …)
•Inflammatory cascade
characterized with
invivo
•Elucidated not only
pathways but cellular
networks (hallmarks).
Define imaging based
on these hallmarks,
leading to go/no-go
•Open access to data
Multi-modality screening
commonly used
•Dose reduced
•Lots of 5mm nodules found
•Backed-up with more
specific test
•Characterize:
•MRS, PET (more
than 1 agent)
•Tell btw e.g.,
squamous,
bronchial cell, etc.
•High specificity
•Array of tracers
that can
differentiate btw
oncogenic events vs.
non
•More widespread
use of data across
care cycle solutions
•Localized, guided
therapy
•Tailored drug
development
•Stratification for disease
based on genotype and/or
mechanistic
•Measures
•Informatics approaches
worked out to be acceptable
workflow
•Multi-targeted
therapeutics
•Signal
amplification
approaches for
targeted for low
SNR