Molecular and Genomic Pathology at UNC

1
Molecular and Genomic Pathology at UNC
2
Molecular and Genomic Pathology at UNC
David A. Eberhard, MD, PhD Dept. of Pathology and
Laboratory Medicine Director, Preclinical Genomic
Pathology Lab Lineberger Comprehensive Cancer
Center University of North Carolina Chapel
Hill ADASP Annual Meeting March 2, 2013
3
Notice of Faculty Disclosure

• In accordance with ACCME guidelines, any
  individual in a position to influence and/or
  control the content of this ASCP CME activity has
  disclosed all relevant financial relationships
  within the past 12 months with commercial
  interests that provide products and/or services
  related to the content of this CME activity.
• The individual below has responded that
  he/she has no relevant financial relationship(s)
  with commercial interest(s) to disclose
• David A. Eberhard, MD, PhD

4
World Rank of NGS Centers
From http//topsequence500.org/
5
Different NGS Platforms Have Different
Capabilities
DNA copy number variations
DNA rearrangements
Sequence alterations DNA and RNA
Text
RNA splicing variants
Methylation
RNA expression profiles
A single method is suitable for some of these,
but not others must consider cost, specimen
type, application
6
NGS Applications in Cancer
Human Genome Project reference genome and
large-scale compilation of tumor variants from
various sources (http//cancercommons.org,
www.icgc.org, http//cancergenome.nih.gov/,
http//www.sanger.ac.uk/genetics/CGP/cosmic/

• Mutation Panels (Genotyping or resequencing)
• Exome or transcriptome screening
• Genome sequencing (compare to normal or
  reference sample)

7
Genome Res. 2012 Nov22(11)2101-8
Erlotinib (2004)
Crizotinib (2011)
Text
Vandetinib
Vemurafinib
Resistance to Erlotinib?
8
Headline
9
NGS For Cancer Diagnostics

• Potentially improve Economy, Efficiency,
  Sensitivity
• No one size fits all must consider desired end
  use
• Considerations for technical platform
• Broad vs Deep More genes vs more sensitivity
• Turnaround time cost single samples vs
  multiplex batches
• PCR vs non-PCR libraries implications for sample
  amount, false positives
• Sample preanalytical variables (FFPE, amount,
  etc)
• References (T/N), standards
• What results does it provide? What results do we
  report?

10
Making NGS Accessible
11
NGS In Clinical Cancer Diagnostics
How much do you need? Broad coverage more
complexity and cost more unknown variants
overkill for clinical care?
What do you need to find? Large indels,
rearrangements with variable breakpoints are
difficult
12
Deep Coverage Improves Mutation Detection
13
Small Sample Size and Low Tumor Content May
Result In False Negatives
Detection Limit
- - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - -
14
NGS Mutation Detection Issues
Mutation confirmation Usually by Sanger
sequencing- will platform evolution eliminate?
NGS platform-dependent false positives
May overlap with NGS false positive rate and
background noise Low level mutations- not
easily confirmed by Sanger sequencing (higher
detection threshold 15-20) May need more
sensitive method DGGE, dHPLC, pyrosequencing
or mutation enrichment- i.e. COLD PCR
FFPE background noise
Variable tumor cells and variable tumor cells
with secondary mutation
15
What to Test?

• Quality and quantity are key determinants
• A cellular FNA is better than a necrotic
  resection
• Decal Bouins etc degrade quality
• Primary vs. metastasis
• May be changes during interval therapy
• If metastasis after initial response, then test
  metastasis
• Multiple primaries
• If histologies differ, then test BOTH/ALL
• Patients benefit even if 1 of multiple tumors
  responds
• Testing multiple areas in a tumor is unnecessary

N Engl J Med. 2012 Mar 8366(10)883-92
16
Minimum Tumor Content

• Absolute and relative amounts of tumor
• Each lab must determine during validation
• Pathologist must review each section
• Enrichment Macrodissection is recommended
• Laser capture, WGA are error-prone

17
Tumor Sample Heterogeneity

• Clinical sample characteristics size/amount,
  matrix, preservation
• Blood whole, buffy coat, Ficoll, FACS, CTCs
• Tissue fresh, fixed (FFPE), decal bone,
  biopsies, resections
• Cytology aspirates (FNA), buccal swabs, smears
• Clinical sample composition Various cell and
  tissue types
• tumor cells, stromal cells, vascular cells,
  immune/inflammatory cells, normal tissue
• Viable tissue (dense, fatty), necrosis, mucin,
  hemorrhage
• Tumor genotypic and phenotypic composition
• Mono- vs Polyclonality tumor evolution variable
  differentiation, EMT, stem cells

18
Bioinformatics
NGS diagnostics is highly dependent on data
analysis and management
Clinical Issues
Evaluation of the variant positions called
involves queries of all known relevant databases
Requires bioinformatics and statistical expertise
and computational hardware
Lack of databases curated to accept clinical
standards is significant challenge in managing
and reporting genome sequencing data
Unprecedented amounts of data and processing
algorithms necessitate adequate tools (Alignment
and assembly QC of image processing, base
calling, filtering, variant calling, SNP finding,
archiving)
EHR considerations test ordering, archiving of
NGS reports, patient consent, data
(reinterpretation?)
19
Clinical Utility - Challenges
Which variants are clinically actionable?
How to establish significance (Structural,
functional, preclinical, clinical)? What are
necessary levels of evidence?
NGS yields many variants of unknown significance
Risk of over interpretation unnecessary medical
action unwarranted psychological stress
20
Headline
Specimen Issues
Predictive Model Development, Specification, and
Preliminary Performance Evaluation
Clinical Trial Design
Assay Issues
Ethical, Legal and Regulatory Issues
21
LCCC 1108 Development of a Tumor
Molecular Analyses Program and Its Use to Support
Treatment Decisions (UNCseqTM)

• Primary Objectives of LCCC1108
• To provide a mechanism for association of known
  molecular alterations with clinical outcome in
  oncology via genetic profiling of patient
  specimens
• To support treatment decisions by providing rapid
  genetic profiling of patient specimens and
  sharing reportable results with treating
  physicians
• Prospective patients are consented such that
  biopsied tissue may be used for both research and
  clinical purposes
• Executive, Technology, Clinical, and Pathology
  committees formed to cover all aspects of the
  study protocol

22
LCCC 1108 (UNCseqTM) Process
23
Targeted Exome Sequencing
Normal
DNA
Libraries
Tumor
UNCseq 6.0 247 cancer genes, 10 viruses
Computational processing to call somatic mutations
pool
Illumina HiSeq or MiSeq
barcode
24
Sequence Alignment
Read
ATGCCATTACACAGCGA
Human Genome (hg19)
CGATCTAACGTAGCTAGCTAGCTAGCTAGCATGCCATTACACAGCGAA
CAGGGAGCTTAGGCGC
GTAGCTAGCTAGCTAGC
GAACAGGGAGCTAAGG
CTAGCTAGCTAGCTAGC
ACAGGGAGCTAAGGCGC
CGATCTAACGTAGCTAGC
ATTACACAGCGAACAGG
25
Tumor Somatic Mutation Calling
26
Glioblastoma Tx resection, chemoradiation. Progre
ssed with transformation to gliosarcoma
Approved drug linked to gene, SOC (1) or Non-SOC
(2A)
Headline
Potential clinical action, e.g. drug in clinical
trials
27
Glioblastoma Tx resection, chemoradiation. Progre
ssed with transformation to gliosarcoma
Approved drug linked to gene, SOC (1) or Non-SOC
(2A)
Headline
Potential clinical action, e.g. drug in clinical
trials
28
Headline
29
UNCseq Gliosarcoma PTEN IHC
Sarcoma component
Glioma component
Did PTEN mutation accompany evolution to sarcoma?
30
Mol/Genomic Path Education UNC

• Molecular Diagnostics course for residents and
  fellows includes 3.5 hrs on NGS
• 2 Mol Path fellows with focus on genomics 1 on
  UNCseq (oncology), 1 on NCGenes (germline)
• LabCorp / UNC interactions provides reference lab
  exposure for Mol Path fellows (diagnostics and
  clinical trials)
• Translational Pathology course for PhD and MD/PhD
  students includes sections on Mol Path, Genomics
  and Translational Pathology Ex-Academia

31
Mol/Genomic Path Education Pharma and Dx
Industry

• Molecular Pathology and Cancer Genomics
  integrates with targeted drug development
  provides tremendous opportunity for cooperation
  between pathology centers and industry
• Providing clinical-grade assays and laboratories
  to support trials and high-quality research
• Providing expertise to integrate practical
  pathology with cutting-edge science
• Expanding educational and career opportunities
  for pathologists

32
UNCseqTM Team

•  Investigators
• Shelley Earp
• Juneko Grilley-Olson
• Neil Hayes
• Ned Sharpless
• Ben Calvo
• Matthew Ewend
• Matthew Nielson
• Linda Van Le
• Robert Esther
• Nirmal Veeramachaneni
• Cary Anders
• Peter Voorhees
• Keith Amos
• Robert Dixon
• Stergios Moschos
• Young Whang
• David Eberhard