Integration of Smart Imaging Technology into Toxicologic Pathology Dr. Gary Knutsen

1
Integration of Smart Imaging Technology into
Toxicologic PathologyDr. Gary Knutsen Robert
LoBosco, Systems Pathology Company, LLC

• Presented to

2
CAPS - Year One
Technology Collaborator
Pathology Collaborators
Status Update 10/22/07

• The Reason for CAPS

• Lessons Learned

• Technology Challenges Applications

• Expectations Changes

3
The Reason for CAPS
SUPPLY
DEMAND
Rapid proliferation of veterinary schools has
stalled and attendance has leveled and in some
cases decreased
Estimated 250 vacant positions in 2007 with more
than 50 of those in industry
Decreased funding (NIH and State Grants) for
advanced training in pathology
Salaries and job versatility demands continue to
escalate as Pharma competes with maturing
biotechs and CROS
Once robust military pathology training programs
are nearly non-existent
Importance of pathology discipline being
recognized with increased numbers assuming key
management/leadership positions and increased
involvement in discovery/early development, and
omics technologies and biomarker research
Pathology programs abroad offer little relief as
compared to other scientific disciplines
European and Japanese Pharma have increased their
State-side recruiting efforts
Percentage of veterinarians pursuing advanced
training is decreasing
US big Pharma and CRO endeavors in Far East are
pulling pathologists from US
Certification (ACVP) examination is challenging
and historical success rates are alarmingly low
Dramatic increase in the number of pathologist
vacancies in Academia and Diagnostic Labs
Challenging demographics - entire generation of
our most experienced toxicologic pathologists
will be retiring in the near term
Salaries and early partnership opportunities for
young veterinarians in clinical practice are
increasingly more lucrative
4
Options for Industry to Consider

• Train more qualified toxicologic pathologists
• ACVP/STP Coalition
• Establish parapathology training programs for
  tissue screening
• - Mega Mouse Project _at_ the National Center for
  Toxicological Research (NCTR)
• - Consideration being given by big Pharma as
  part of Far East RD expansions
• - Significant validation and Regulatory
  challenges
• Computerized Automation (Smart Imaging) of
  Tissue Screening
• Significant advances in digital imaging
  platforms and supporting technologies
• Automation success in other biomedical
  applications
• Automated Paps Testing
• Computer Assisted Diagnosis (CAD) in Radiology
• Automated Cell- and Molecular-based assays
• Virtual microscopy solutions for remote image
  assessment and automated morphometry
• CAPS showing promise as a software imaging
  tool to augment toxicologic pathology
• Projected increased productivity and
  through-put efficiencies
• System objectivity augments subjective
  objectivity of pathologists

Todays technology makes CAPS the most immediate
and viable option
5
The Parts Existed but were not Aligned
Hardware Software
No one organizational entity has all the parts
necessary for CAPS Development Success ---
Effective collaboration is a must!
6
The SPC CAPS Development Process
Specimen Access
Specimen retrieval with supporting SPC
Toxicologic Pathologists
Federal Agencies, Academia Institutes
End User Participation
7
Pathology Annotation Workflow
Local Network
SPC Network
8
The Future with CAPSFully Integrated Workflow
LIMS Study Data
Necropsy
Histology
Aperio ScanScope XT-S
9
HistoQC
Automated screening of slides for histologic
quality and monitoring samples for Automated
Toxicologic Pathology (HistoPATH)
Anatomy Verification

• Verifies adequate anatomy represented in all
  tissue samples
• Incomplete sections are rejected

10
Automated Toxicologic Pathology HistoPATH

• Retrieves Historical Normal Controls to define
  Normal for study
• Evaluates Study Controls
• NormalGreen, Non-normalYellow
• Evaluates Study Treated animals
• NormalGreen, Non-normalRed

11
HistoPATH - Smart Imaging
But, does the software know its a Kidney?
or whether its a Normal or Non-normal Kidney?
HistoPATH automatically recognizes and analyzes
all protocol-specified tissues in a study,
separating the Normal from the Non-Normal
(gloms too!)
12
HistoPATH Development
Our primary focus this past year has been on
HistoQC
but the technology applies directly to HistoPATH
The same software that detects this tiny Floater
will also detectthis.
13
Sponsor Proof of Concept
Client Re-exam Assessment
SPC Pathologists
Client Pathologists
Rat Liver Group
HistoPATH
Group A 35 Historical Normal Controls
Group B 15 Controls, 45 Treated
Group C (Blinded) 20 Controls, 20 Treated
CAPS identifies and will help lower Objective
Subjectivity
14
Anticipated Efficiencies
Estimated 7 full days of actual histopathology
evaluation - including entry in DMS
Traditional Histopathology Approach

• Day 0
• Overnight Automated Analysis
• HistoQC Screening
• Normal v. Non-normal Analysis

• End of Week 1
• Evaluation completed
• Report preparation initiated

• Pathologists Time CAPS (2) days v. Traditional
  (7) days.
• CAPS Report 1 week v. Traditional 3 weeks

HistoPATH Approach
Week 1
Week 2
Week 3
15
A High Content Toxicologic Pathology Environment
LIMS Study Data
More Accurate Predictive Toxicology
Improved Translational Medicine
Necropsy
Histology
Aperio ScanScope XT-S
16
Regulatory Compliance Considerations

• The software development and documentation
• Repeatable and measurable processes
• ISO 90012000
• 21 CFR 11 (Electronic Records/Signatures) and 820
  (Medical Devices)
• 21 CFR 58 (Good Laboratory Practices)
• FDAs General Principles of Software Validation. 
• Quality Management System
• Rigidly responsive to requirements and applicable
  standards
• Subject to constant internal audit/monitoring and
  routine external audit
• Performance Testing and Validation
• SPC/TPA Internal Performance Testing
• FDA/NCTR Collaboration Peer Review/PWG
  Validation
• User Validation
• Concurrent Traditional vs. CAPS Assessments
• Conformance Testing
• Proposed as being analogous to Automated Clinical
  Chemistry Validation
• Validated Tissue Images analyzed with each study

CAPS development accommodates an ongoing
Regulatory Compliance process
17
Before Smart Imaging
18
After Smart Imaging