Transcript Establishing a Clinical Genomics Service at an Academic Medical

Establishing a Clinical Genomics
Service at an Academic Medical Center
Jason Merker, MD, PhD
Co-Director, Stanford Medicine Clinical Genomics Service
Regional APC Meeting, Maui
October 25-28, 2016
Outline
I.
II.
III.
IV.
Background on clinical genomics
Workflow and case example
Four major challenges
Update and future directions
Background
Clinical Genomics
Wellcome Collection – Medicine Now
3.4 billion units of DNA code:
• 127 volumes
• 1,000 pages per volume
Size of targeted regions in assays
Genome – 127 volumes
Exome – 2 volumes
400 gene heritable CA panel –
90 pages
CF hotspot panel – < ½ page
Single gene – 1-2 lines
Amount of data generated for a single
exome at 150X = 3 bookcases
Genome
Exome – 300 volumes
400 gene heritable CA panel
CF hotspot panel
Single gene
Amount of data generated for trio genome
analysis ≅ 135 stacked bookcases
Honokohau Falls,
West Maui
1,119 ft (341 m) =
~100 bookcases
Decreased sequencing costs with NGS
Sequencing centers
transition to NGS
Desktop sequencers
Growth of DNA Sequencing
2016 – 1,800 genomes/instrument/yr
2006 – 1 genome/instrument/yr
Stephens ZD et al. (2015) PLoS Biol 13(7): e1002195. doi:10.1371/journal.pbio.1002195
Workflow and case example
Stanford Medicine Clinical Genomics Service
Stanford Medicine Clinical Genomics Service
• Pilot phase started in January 2014
• Use genome and exome sequencing
to determine the cause of disease in
patients with suspected genetic
disorders
• Focused on three major disease
areas:
– Pediatric and adult syndromes
– Heritable cancer predisposition
– Heritable cardiovascular disease
Workflow for Clinical Sequencing
1 – Test ordering & review
Genetic Counselor
Biocurators/informatician
Molecular
Geneticists/Pathologist
Test request by
Stanford health
care provider
Utilization
review
Treating Team
Analysis
Team
Clinical Genetics Expert
Establish questions being
posed by patient and
treating team
Workflow for Clinical Sequencing
2 – Counseling, consent & specimen collection
Insurance
authorization
Clinical counseling & consent
• Molecular diagnostic yield
• Incidental/Secondary findings
• Biobanking and data sharing
Specimen collection
• Genome/exome
sequencing
• Specimen
ID/confirmatory
studies
• Biobanking*
Workflow for Clinical Sequencing
3 – Sequencing and data analysis
Pipeline Definition – CWL/YAML File
LOOM – Run
Anywhere,
Reproducible
Workflows
Genome or exome
sequencing
(Illumina chemistry)
Variants
Data analysis (MedGAP)
• Alignment
• SNV/indel calling
• QM program
• ID checks to confirm
specimen identity
Annotation,
filtration &
prioritization
• Phenotype
• Variant
impact
• Inheritance
• Pop freq
• ClinVar
Workflow for Clinical Sequencing
4 – Verification and report generation


Genomics Review Groups
• Genomics Service
• Treating Team
• Clinical genetics expert
Variant verification
by orthogonal
method
• Segregation
analysis
Curation meeting & draft
variant classification
• Genetic counselors
• Clinical data scientists
• Bioinformaticians
• Molecular
pathologist/geneticist
s




Genomics Review Board
• Additional genetics
expertise
• Bioethics/legal
Workflow for Clinical Sequencing
5 – Reporting & post-test counseling
Generate final
report
• Coverage
metrics of
relevant genes
Patient meets with genetic
counselor and relevant
members of treatment
team
Reanalysis
upon request
Case example – one family at a time…
Two affected siblings with:
• Progressive cerebellar
ataxia
• Progressive atrophy of
the cerebellum
• Intellectual disability
• Decreased muscle tone
and reflexes
• Delayed development
• Distinct facial features
Findings suspicious but not conclusive
enough to act on clinically
• We performed genome
sequencing on this family and
found that both copies of a
gene (SNX14) were inactivated
in the affected children
• Studies in mice indicated that
SNX14 is involved in normal
brain and nervous system
development
• BUT, this gene had not been
implicated in human disease
Our service and the treating physicians believe
this is the cause of disease in this family
• Ends an ~10 year diagnostic odyssey
• Provides family information about the
disease course and associated management
approaches based on the experience of >30
patients
• This gene is involved in a cellular process for
which there is a therapeutic target that is
under investigation in other
neurodegenerative disorders
Percentage of “solved” cases
Patient population
Pediatric syndromes
Heritable cardiovascular
disorders
Heritable cancer
predisposition
Primary test
31%
Secondary test
29%
35%
17%
40%
7%
Molecular diagnostic yield = Pathogenic (P) + likely pathogenic (LP) variants
Four major challenges
Stanford Medicine Clinical Genomics Service
Challenge 1: Personnel recruitment/retention
Faculty
• Directors
(Ashley/Merker)
• Clinical Molecular
Geneticist
Bioinformatics
• Director of
Bioinformatics
• Bioinformaticians (3)
• Software Engineers (2)
• Testing Engineer
Laboratory
• Supervisor
• Clinical Laboratory
Scientists (2)
Interpretation
• Clinical Data Scientists
(2)
• Genetic Counselors (2)
Administration
• Administrative Director
• Administrative Assistant
Challenge 2: Organization and coordination
Stanford Medicine Advisory Committee for Genomics
Name
Organization/Department
Role
C. Dawes
Stanford Children’s Health
President and CEO
L. Minor
School of Medicine
Dean
D. Entwistle
Stanford Health Care
President and CEO
T. Quertermous*
CV Med
GenePool Biobank
J. Ford*
Oncology
Chief Cancer Genetics
E. Ashley
CV Med
Institute for Inherited CV Disease
L. Boxer
Hematology
Vice Dean
M. Cho
Pediatrics
Center for Biomedical Ethics
L. Hudgins
Pediatrics
Chief Medical Genetics
M. Leonard
Pediatrics
Chair Pediatrics, Physician in Chief SCH
J. Merker
Pathology
Co-director Clinical Genomics Service
T. Montine
Pathology
Chair Pathology
K. Ormond
Genetics
Genetics and Genomics Counseling Program
M. Snyder
Genetics
Chair Genetics, Director SCGPM
D. Stevenson
Pediatrics
Senior Associate Dean Maternal and Child Health
Challenge 3: Compute and storage
Google Cloud Platform
• Limited ability to expand on premise
• Build for scale with an uncertain ramp
• Find solution that meets security, legal and
other business requirement of three distinct
entities
• HIPAA-compliant implementation
Challenge 4: Software and systems
Director of Bioinformatics and Team
Adult
EPIC Beaker LIS
Genomics Information System
- Workflow manager
- Analysis pipeline
- Variant database
Children
Update and future directions
Stanford Medicine Clinical Genomics Service
Transition from translational pilot to
clinical service around September 2017
Dry Laboratory
Moved in September 2016
Cloud-based Compute & Storage
Dev. environment - ETC November 2016
Prod. environment – ETC September 2017
Wet Laboratory
ETC December 2016
Other (in silico) panels:
• Heritable cardiovascular
• Heritable CA predisposition
• Neurologic disease
What’s next…
• Improved sequencing approaches
– Fill in coverage gaps
– Longer read length to identify difficult to call variants
– Incorporation of ancillary data (e.g., RNA-Seq)
• Improved data analysis
– Identification of difficult to call variants (larger indels, CNVs, SVs)
– Analysis of non-coding variants
• Increased gene-disease and variant-disease associations
– Clinical findings coded with standard terminology
– Data sharing efforts
• More efficient and improved counseling, consent and result
return
• Better incorporation of genetic data into clinical practice
– Provider education
– EMR integration
Acknowledgements
Stanford Health Care
Clinical Laboratories &
David Entwistle*
Pathology
Stanford Medicine Advisory Thomas Montine*
Amir Rubin
Clinical Genomics Service Committee on Genomics*
James Hereford
Stephen Galli
James Ford
Euan Ashley*
Pravene Nath
Merrie Bass
Thomas Quertermous
Somalee Datta
Stephen Ayers
Tena Cherry
Linda Boxer
Dianna Fisk
Tom Bruynell
John Christopher
Mildred Cho
Jim Ford
David Connor
Manijeh Danaye-Elmi
Louanne Hudgins
José González
Gary Fritz
Shirley Weber
Mary Leonard
Megan Grove
Monica Gupta
James Zehnder
Kelly Ormond
Nathan Hammond
Tarun Mahendroo
Michael Snyder
Louanne Hudgins
Stanford Children’s Health Jean-Raymond X.
David Stevenson
Ruchi Joshi
Pierre
Christopher Dawes*
Isaac Liao
Balaji Ramadoss
Anne McCune
Stanford University SOM
Zena Ng
Suzanne Roosevelt
Ed Kopetsky
Dean
Lloyd
Minor*
Ranjan Muthumalai
Neil Shah
José González
Michael
Halaas
Suzanne Roosevelt
Christopher Sharp
Mary Leonard
Carlos Suarez
Mohan Vasan
Harriet Nibbelin
Tam Sneddon
Yohan Vetteth
Natalie Pageler
Sowmi Utiramerur
Steven Wright
Mark O'Connor
Shana White
Melody Zhang
https://stanfordhealthcare.org/clinicalgenomics
Amin Zia
SHC and SCH patients and health care teams