Transcript IBM Healthcare and Life Sciences

IBM Healthcare & Life Sciences
Clinical Genomics:
A Path Towards Information-Based
Medicine
A New Era in Patient Care
Kareem M. Saad
WW Business Segment Executive, Clinical Genomics
Information Based Medicine
IBM Healthcare and Life Sciences
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Agenda
Information Based Medicine: A New Era in Patient Care
IBM Healthcare & Life Sciences Clinical Genomics Solution
Application of Clinical Genomics Solution
Implementations of Clinical Genomics Solution
Page 2
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Emerging Business Opportunities (EBO’s) at IBM
Somehow change dynamics in the marketplace
 shift in business model
 new set of customer requirements
 disruptive technology
Don’t represent business as usual
 need care and feeding
Create infrastructure around each EBO
 Test, explore, or to morph into another opportunity
Page 3
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
The Information Based Medicine EBO Growth
Currently 1 of only 5 EBO’s across IBM
Criteria Across EBO Life Cycle
US$ 1 Billion+
Revenues
Selection Criteria
Graduation Criteria
• US$ 1 billion+ potential
• Strong leadership team
in place
• Cross–business
strategic importance
• Clearly articulated
strategy for profit
contribution
• Market leadership
potential
• Market success
• Support from business
unit management
Selection
Page 4
• Proven customer value
proposition
Cultivation
Graduation
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Healthcare and pharmaceuticals are facing
unprecedented challenges driven by market forces
Healthcare Drivers
Aging and informed population
Pharma Drivers
Lack of R&D Productivity increase
despite sustained investments
Demand for safety and efficacy
Page 5
Cost Pressure on prescription
drugs that don’t always work
Blockbuster model no longer
assuring adequate sales and profit
growth
Increased focus on preventive
care / wellness
Unacceptable failure rates of R&D
projects during pre-clinical and
clinical development phases
Progress in science and
technology
Serious drug side effects threatening
Pharma industry
Advances in information
technology
Need to leverage industry knowledge
in patient care / disease mgmt
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Resulting in transformative changes to the delivery of
and quality of healthcare
Genetic testing routine for some population groups
Many major diseases understood at molecular level
Subpopulations at risk for adverse drug events will be identified for
many therapeutics
Companion diagnostics developed with targeted therapeutics
Increased transition towards wellness care
Page 6
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Revolutionary Technology
Automated
Systems
Information
Correlation
1st Generation
Diagnosis
Distributed High-Throughput Analytics
This transformation is being accelerated by a combination of
revolutionary technologies and evolutionary practices
Personalized Health Care
Lifetime Treatment
Pre-symptomatic Treatment
CA Diagnosis
Translational
Medicine
Molecular Medicine
Genetic Predisposition Testing
Clinical Genomics
Health Care
Digital Imaging
Today
Episodic Treatment
Electronic Health Records
Artificial Expert Systems
Data and Systems Integration
Non-specific
(Treat Symptoms)
Organized
(Error Reduction)
Personalized
(Disease Prevention)
Evolutionary Practices
Page 7
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
The Future of Healthcare Lies in “Information
Based Medicine”
Utilizing Information Technology to Achieve Personalized Health Care
“The future is not so hard to predict. It’s already here. It’s just not equally distributed.”
William Gibson
New Treatment
Options
Life Sciences
Legacy
Pharma
Pharma
R&D
Healthcare
Pharma
Pharma
Delivery
Delivery
“InformationBased Medicine”
Clinical
Data for
Research
New drugs
delivered
Page 8
Clinical
Decision
Support
Legacy
Healthcare
Improved patient
safety and quality of
care
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Information Based Medicine will require unprecedented
access to diverse, integrated information
Challenges
• Volume and complexity of data
• Integrating massive volumes of
disparate data
•Need for sophisticated analytics
•Growing collaboration across ecosystem
1. Patient
Information
Hospital events ....admission,
surgery, recovery, discharge
Access to Diverse
Heterogeneous
Distributed Data
Expression Arrays
(various tissues)
Personal
genomics
X-rays, MRI,
mamograms,
etc
Clinical Record
Page 9
Analysis
lab notes
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Academic Medical Research Centers & “Biobank” Challenges
•
Medical
Research
Care
Delivery
Patient
•
•
•
Academic Medical Research
Centers (AMRC’s) and
Biobanks focus heavily on the
integration of care delivery and
research functions to improve
understanding of disease,
improve quality of care and
ultimately move towards
personalized medicine
Page 10
•
•
•
Collect, access or mine all patient data
(including blood and tissue samples)
securely for research
Patient records not well organized –
difficult to find patient history/need to build
enterprise clinical data warehouses
Compliance with regulatory requirements,
particularly patient privacy regulations
(e.g. HIPAA in US)
Emerging role of genomics requires patient
records to be enhanced with new
molecular profiling data
Leverage data standards to enable
linkages and associations between
disparate data types (e.g. phenotype and
genotype)
IT departments not equipped to deal with
complexities of the data integration
challenge faced by translational research
Establish leadership position to attract gov’t
and industry funding as well as attract top
medical researchers
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Biopharmaceutical R&D Challenges
•
Pharma
Discovery
Clinical
•
Development
•
Bio-Pharmaceutical R&D
organizations are adding
targeted treatment capabilities
requiring investment in new
technologies and IT
infrastructures for blood and
tissue sample management and
patient databases that
incorporate genotypic data, and
query, analysis, and mining
capabilities.
Page 11
•
•
•
•
•
Identify and validate specific drug targets
associated with chosen therapeutic areas
Collect, access or mine patient data
(including blood and tissue samples)
securely for research
Recruit patient populations characterized
by Biomarkers identified by diagnostic tests
“Rescue” drugs that failed due to side
effects in subsets of patient populations
characterized by shared Biomarkers
Manage blood and tissue samples
collected during clinical trials
Compliance with regulatory requirements,
particularly patient privacy regulations
(e.g. HIPAA in US)
Leverage data standards to enable
linkages and associations between
disparate data types (e.g. phenotype and
genotype)
Pharma IT departments must deal with
complexities of the data integration
challenge faced by need to combine
patient data with genomics, biomedical
image and literature data
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Agenda
Information Based Medicine: A New Era in Patient Care
IBM Healthcare & Life Sciences Clinical Genomics Solution
Application of Clinical Genomics Solution
Implementations of Clinical Genomics Solution
Page 12
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Clinical Genomics – What is it?
“Clinical genomics is the marriage of large scale
technologies for molecular analysis with the study of
actual disease.”
-Cambridge Healthtech Institute
Different names intended to
describe the same thing:
•Clinical Genomics
•Medical Informatics
•Pharmacogenomics
•Personalized Medicine
•Healthcare Informatics
•Translational Medicine
•Medical Genomics or
Genomic Medicine
•… and on and on!!!
Page 13
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Clinical Genomics: Conceptual Overview
The
Thetechnological
technologicalchallenges
challengesbehind
behindthe
thecollection,
collection,storage,
storage,management,
management,
integration
integrationand
andanalysis
analysisof
ofphenotype
phenotypeand
andgenotype
genotypedata
datafor…
for…
Phenotype:
Genotype:
• EMR / EDC Data
• Patient outcome data
• Cell Image (e.g. IHC)
• Medical Images
• Disease Progression
• etc…
• Raw sequence
• SNP’s/Microsatellites
• Transcriptome
• Proteome
• Genealogy
• etc…
1. Identifying and validating novel therapeutic targets
2. Conducting more focused clinical research (Pharmacogenomics)
3. Revolutionizing the ways by which diseases are diagnosed and treated
Page 14
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
IBM Healthcare and Life Sciences Clinical
Genomics Solution Conceptual Architecture
Medical
Research
Expression, SNPs,
Clinical Studies &
Trials, Proteomic
Clinical
Care
HIS, RIS, CIS,
Pathology, Rx,
Patient Charts
Data Mining/Statistical
Analysis/Visualization
Adherence to Standards
HL7, BSML/HapMap, CDISC/ODM,
MAGE-ML, CDA, etc.
Medical Information Gateway
Deidentification of Patient Data & Anonymous
Global Patient Identifier Assigned
Medical Information Broker
Page 15
WebSphere
DB2 Information Integrator
Medical
Information
Repository
Source scientific data &
unstructured text files
e.g. MS Access. MS Excel,
EST/ GenBank, XML,
Medline, dbSNP
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
The Clinical Genomics Solution enables multiple institutions to
seamlessly and securely share information for collaborative
research and improved clinical care
MIG
MIB
Data Mining/Statistical
Analysis/Visualization
WebSphere
Page 16
MIG
MIB
MIG
MIB
DB2 Information Integrator
Medical
Information
Repository
Source scientific data &
unstructured text files
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Anonymous Global
Patient ID
…
Web Services
deCode CGM-D
Data
Handlers
Micro Strategy
CDA
Constructor
SpotFire
AGPI Server
Data
Handler
SAS
Optional Applications
Search, Query, Mining,
..
…
Open Source
Researchers
DDQB
Medical
Information
Administration
Service
SNP
Clinical Studies
Gene
Expression
Patient Charts
CDA
HL7 2.x
Interface
Engine
CDISC / ODM
MAGE-ML
CG V2 Architecture
BSML / HapMap
HIS, CIS, RIS, Pathology, Rx
WebSphere
De-Identification
DB2 II
CDA, MAGE-ML,
BSML / HapMap,
ODM
Page 17
Relational
Connect
PubMed
Flat File / XML
BLAST
Medical
Information
Repository
DB2
ODBC
Medical
Information
Broker
Parser /
Loader
Correlation / Curation
Medical Information Gateway
MS Access / Excel
EST / GenBank
Delimited / XML Files
Medline
dbSNP
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Medical Information Gateway (MIG)
Functionality
Phase I:
Enablers &
Inhibitors
Medical
Information
Gateway
Page 18
• Interfaces with systems in the hospital and research settings to
capture, normalize, de-identify and aggregate information based
on each patient or research encounter
• Captures diagnostic, clinical, demographic, gene expression,
clinical study, phenotypic, life style or environmental data from a
range of lab or hospital systems (e.g. CIS, HIS, RIS)
•Transforms information into the appropriate data standards, deidentifies the data according to local regulatory requirements and
assigns an anonymous global patient identifier
Benefits
• Securely de-identifies patient data to ensure adherence to
regulatory requirements (customizable using administrative
interface)
• Multiple Gateway’s can be deployed at disparate sites or
departments, enabling secure, cross-institutional collaboration
• Ability to interface with multiple existing hospital and research
systems ensures that existing IT investments are leveraged
• Adheres to industry and open standards ensuring future flexibility
and extensibility of the integrated research and clinical
environment
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Medical Information Repository (MIR)
Phase I:
Enablers &
Inhibitors
Medical
Information
Gateway
Page 19
Functionality
• Integrated data store of all the phenotype and genotype
information captured by the Medical Information Gateway and
delivered through the Medical Information Broker
• Solution can be deployed with a single MIR serving as a large
data warehouse or with multiple MIRs accessible through DB2
Information Integrator’s federated technology
• Leverages IBM’s robust and secure DB2 technology
Benefits
• Accelerates understanding of complex diseases at the molecular
level by providing an integrated environment for the analysis of
large sets of genotypic and phenotypic information
•Multiple third-party, open source or proprietary data mining,
visualization and analytics can be used to query and mine the
MIR due to adherence to open and industry standards
• Links clinical and research environments seamlessly and
enables secure information and secure data sharing across
departments and institutions
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
DB2 Information Integrator
Search
Find
Federate
SQL
Content
Transform
Place
Publish
DB2
Information
Integrator
Shared
metadata and access
foundation
Phase I:
<XML>
Text
</XML>
Enablers &
Inhibitors
Functionality
DB2
Information
Integrator
• Real-time, integrated access to business information –
structured and unstructured, mainframe and distributed, public
and private, IBM and non-IBM – across and beyond the
enterprise
• Clinical Genomics-specific support: access structured scientific
data and unstructured text files via federation, replication,
triggered data publication, and enterprise search capabilities
Benefits
• Integrate information across multiple business processes
• Get more value and insight from existing assets
• Control IT costs by tailoring views and reducing
Any data
redundancy
Multiple access paradigms
Multiple integration disciplines
Page 20
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
IBM Healthcare and Life Sciences large ecosystem of partners*
across the clinical and research domain, helps to ensure
seamless integration and minimal disruption to operations
* NOTE: Not a
comprehensive list
Page 21
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Agenda
Information Based Medicine: A New Era in Patient Care
IBM Healthcare & Life Sciences Clinical Genomics Solution
Application of Clinical Genomics Solution
Implementations of Clinical Genomics Solution
Page 22
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
A major University Health System consists of the primary
University Hospital, 5 affiliated regional hospitals and a large
cancer center
University Hospital
• 400 bed hospital w/ over 20K patients
per year
• Clinical and imaging systems not
integrated
• Participation in multiple clinical trials
Affiliated Hospital Centers
• 5 affiliated health centers
• Clinical systems loosely integrated
with University Hospital
• Participation in multiple clinical trials
sponsored by University
University Cancer Center
• University Cancer Center affiliated
with Hospital
• Large genotyping and molecular
profiling capabilities
• Receives substantial government
funding for basic research
Page 23
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
The University Health System implemented the Clinical Genomics
Solution to integrate genotypic, phenotypic, clinical trials and
patient data from each hospital and research site in network
Data Types
Clinical trials
(ODM), Clinical
(ODM), Imaging
MIG
MIB
Data Mining/Statistical
Analysis/Visualization
De-identification
WebSphere
Data Types
Expression
(MAGE), Mutation
(BSML), Genotype
(BSML)
MIG
MIB
DB2 Information Integrator
De-identification
Data Types
Clinical trials
(ODM), Clinical
(ODM), Imaging
MIG
De-identification
Page 24
MIB
Medical
Information
Repository
Source scientific data &
unstructured text files
e.g. dbSNP, PubMed,
GenBank
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
The integrated system enabled collaboration between a now
networked clinical and research environment
Scenario
 Hospital has been running a large, pharmaceutical sponsored clinical trial for lung
cancer
 Mechanism of action of the compound is not precisely known and a limited sub-set
of patients are showing efficacy
 Several patients enrolled in study are experiencing side effects
 An oncologist involved in the trial has discussed this with an oncology fellow and
asked the fellow to investigate if the adverse effects or compound mechanism of
action can be identified
Data Types Leveraged in Scenario
 Clinical Data (CDA)
 Gene Expression (MAGE)
 Clinical Trial (ODM)
 Mutation Data (BSML)
 Genotype (HapMap)
 Unstructured Text (PubMed)
 dbSNP
Page 25
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Research fellow queries de-identified genomic and clinical
information in MIR to identify all lung cancer patients at network of
medical centers
Deidentified,
anonymous
global patient
identifier
Page 26
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Research fellow able to identify lung cancer patient of interest
without revealing identity because of deidentification
Page 27
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Fellow is able to query PubMed through same interface as
MIR queries leveraging DB2 Information Integrator
Paper indicates poor survival rates for patients with lung
tumors containing K-ras mutations.
Page 28
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Oncology fellow queries clinical, gene expression &
mutation information simultaneously in MIR for patient X
Fellow is able to determine that patient
X shows SNP at amino acid 12.
Page 29
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Leveraging DB2II, the fellow is able to seamlessly query dbSNP to
determine if the mutation is already reported in the public DB
Location of SNP for K-ras 2 protein in dbSNP is at the 161 amino
acid position. Patient X shows SNP at amino acid position 12 – this
could be a publishable finding.
Page 30
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Fellow then queries MIR to determine if patient x is enrolled
in any clinical trials at the medical center
Fellow is able to determine that the trial includes chemotherapy
agent Doxorubicin Hydrochloride and Carboplatin.
Page 31
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Fellow then queries clinical trials data for vital signs and
patient information in the trial
Fellow is able to examine patient’s vital signs as well as tumor
information. Determines tumor is moderately differentiated.
Page 32
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Oncology fellow then queries clinical information in Medical
Information Repository to see deidentified clinical history for
patient X
Patient X clinical history shows that he has
history of heart problems.
Page 33
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Levering DB2 II, the fellow queries PubMed recognizing that due
to patient X’s history of heart disease attention must be given to
hypersensitivity reactions exhibited by anti-tumor agents
Paper indicates that Carboplatin has a role in coronary vasospasm.
The paper concludes that cancer patients collapse when given 450mg
or more of Carboplatin to treat their tumors.
Page 34
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Fellow queries MIR to access clinical trial data to determine
amount of Carboplatin administered to patient X
Fellow is able to determine that level of Carboplatin administered
during trial is likely non-toxic.
Page 35
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Fellow queries for another patient in the clinical trial who
has been given a similar dosage as patient X
Page 36
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Fellow queries for patient Y’s clinical phenotype information
to compare with patient X.
Patient Y, who is enrolled in the same trial and has been
administered similar dosage in trial, does not show signs
of heart disease like patient X.
Page 37
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Fellow queries MIR for genotypic information on patient X
and patient Y
Patient X
Patient Y
He predicts that the phenotypic differences seen between these 2
patients who are on the same drug trial could possibly be due to the
genotypic difference.
Page 38
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Agenda
Information Based Medicine: A New Era in Patient Care
IBM Healthcare & Life Sciences Clinical Genomics Solution
Application of Clinical Genomics Solution
Implementations of Clinical Genomics Solution
Page 39
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Recent Clinical Genomics Projects:
University of California San Francisco
Challenge:
Accelerate development of pharmacogenomic treatments by integrating diverse
data in clinical research.
Focus: Neurology, Cardiology and Oncology.
Solution:
A Clinical Genomics Information Mining system to facilitate the simultaneous
mining/query/analysis/ and cross association of all of the data generated by the
clinician researcher and the associated clinical data from the hospital.
Benefits:
The system will depict associations within the clinical and research data that
helps the researcher formulate or explore new hypotheses of the genetic and
biological basis of disease. This will enable more accurate diagnoses and
tailored treatment, resulting in improved patient outcomes.
Page 40
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
University of California, San Francisco (UCSF)
Differential Diagnosis of Dementia - Challenge
“Gold
standard” for
diagnosis =
Pathology
Patient
Dementia diagnosis
Alzheimer’s
Rapidly-progressing
dementia
Non Rapidlyprogressing dementia
Is there a combination of clinical, pathological, molecular and submolecular “markers” that would allow the better and earlier
diagnosis of various forms of dementia?
Page 41
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
University of California, San Francisco (UCSF) Differential
Diagnosis of Dementia - Architecture
Hospital
Information
Systems
Medi cal Recor d
Pat hol ogy
Labor at or y
Clinician
Clinician
Clinician
Clinician
Scientist
Scientist
Scientist
Scientist
Technician
Technician
Demogr aphi cs
Novel markers for
differentially
diagnosing dementia
MRI
Cogni t i ve Test s
Phar macy
Quest i onnai r es
Information
Information
Integration
Integration
••Data
DataWarehousing
Warehousing
••Data
Federation
Data Federation
Data
Data
Analysis
Analysis
Ani mal Model
Physi ol ogy
I HC
Student
Student
Mi cr oar r ay
SNP Genotypes
Page 42
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
University of California, San Francisco (UCSF)
Long-term Strategy
Disease Area
Multiple Sclerosis
Dementia
Clinical Development
Discovery
Institution #1
Institution #2
Functional Focus
(as part of the
information-based
medicine continuum)
Institutions
(internal and external
stakeholders)
Page 43
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
H. Lee Moffitt Cancer Center and Research Institute
Background
Largest specialty tertiary cancer center in Florida (30% of all
cancer patients in the state of Florida covered by Moffitt
Network)
NCI designated Comprehensive Cancer Center
 Patient Care
 Clinical and Basic Research
 Education (residency programs and affiliation to USF)
In-patient functions, out-patient functions, SE largest Bone
Marrow and Transplant program and Lifetime Cancer
Screening
Mandate “Translational Research”: focused on the rapid
translation of scientific discoveries to better patient care
Page 44
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
H. Lee Moffitt Cancer Center and Research Institute
Problem Being Faced
Lung
Cancer
patients
A single treatment regimen of
both chemo and radiotherapy
Responders
Non-Responders
Lung
Cancer
patients
Differential diagnosis based
on known AND unknown
disease markers
Page 45
Responders
Treatment
Regimen #1
Responders
Treatment
Regimen #2
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
H. Lee Moffitt Cancer Center and Research Institute
Solution
Hospital
Information
Systems
Clinician
Clinician
Scientist
Scientist
Technician
Page 46
Electronic Medical
Record
FDA Submission
Pathology
Laboratory
Case Report
Forms
Research Data
•Gene Exp
•Sequence
•Proteomics
Information
Integration
• Data Warehousing
• Data Federation
Data
Analysis
• Open Standard Representation
• Regulatory Compliant Implementation
Feedback into
clinical development
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
IBM/Mayo Clinic Clinical Genomics Collaboration
Phase I: Example of a real-time query
Find all patients with:








Coronary artery disease (a form of heart disease)
Diabetes Mellitus (“diabetes”)
Nonalcoholic steatohepatitis (a form of liver disease)
Who had a breast biopsy at Mayo (a procedure)
In ZIP code 55901, 55902, 55903, 55904 (local region)
Between 45 and 65 years of age (certain age)
Who are female (female gender)
And are alive (vital status)
Before
After
Page 47
6 Weeks
6 Seconds
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
IBM/Mayo Clinic Collaboration
Applied Genomics Data Analysis
Genomic data (DNA) – GeneChip array data (RNA)
Protein data
Clinical Data
Signs
Symptoms
Laboratory
Radiology
Etc.
Phase I
Page 48
Databases
Genome
Proteome
Disease
Tumors
Drugs
Optimized, individualized healthcare
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
During Phase I, the Mayo Clinic Partnership has produced
one of the world’s largest Clinical Data Warehouses
Warehouse contains over 4.4 Million Patient Records
Infinite number of unique queries across
28 demographic elements
523 DRG codes
10,455 ICD-9 codes
All structured laboratory test conditions or results (up to 4900+)
All microbiology organisms by name; heart rate on ECG
Mayo researcher benefit – months to minutes time
savings for select cases tested
Page 49
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
In Phase II, Mayo and IBM will focus on Genomic Data and
Text Analysis
 Part 1 - Storage and Retrieval of Genomic data
 Genomic data storage and retrieval utilities incorporated into
the data warehouse
 Part 2 - Genomic Data Analysis Workflow
 Genomic data workflow encompassing DNA/RNA test results,
analysis of raw data (e.g., microarray), with
annotation/comparison to reference databases and inclusion
in study list prototype application
 Part 3 - Text Analysis
 Concept-based inquiry and retrieval of unstructured data in
Clinical Notes and Laboratory Reports
Page 50
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Example of Query
… able to run at end of Phase II
“Find all living female patients with diabetes with a
good quality microarray experiment”
 Diabetes Mellitus (Diagnosis Codes, Medical Index & Clinical Notes)
 Serum Glucose > 150 mg/dL (Results)
 Microarray data exist (Storage & Retrieval)

cRNA labeling efficiency > 90% (Workflow)
 Between 45 and 65 years of age at first diagnosis (Demographics
combined with Diagnosis Codes, Medical Index & Clinical Notes)
 Who are female and alive (Demographics)
Page 51
IBM and Mayo Confidential
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Phase II Solution Architecture
DDQB
HSR
SAS
R
S-PLUS
Spotfire
Bioconductor
Affymetrix
Genes@Work
Laboratory Results
and Reports
Registration, RAS,
Diagnoses Codes
and Medical Index
Clinical Notes
Gene Expression
Gene Sequences
Length Polymorphs
Researchers
Search, Analyze,
Export Data and
Results
Web Services
Page 52
DB2
Document
Driver
Annotators
Standard or New XML format
Shredders
MAGE-ML
Standard or New XML format
DDQB reference data
JEDII
CDA / XML
Replication
GoldenGate
ETL
XML transform
XML transform
XM transform
XML transform
Websphere
Life
Sciences
Warehouse
CAS
Extraction
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
IBM’s Commitment to Healthcare and Life Sciences
IBM Healthcare and Life Sciences
 Aligned Clinical Environment
 Healthcare Collaborative Network
 Patient Centric Healthcare Portal
 Clinical Decision Intelligence
 Clinical genomics
 Safe and Lean Hospital
IBM Business Consulting Services
 Component Business Model
 IT/business consulting
IBM Partners who provide Applications





Clinical and Enterprise Information Systems
Patient Accounting Systems
Admitting Systems
Patient Care Systems
Ancillary Systems: Laboratory, Pharmacy,
Radiology
 Physician Office Systems
 Life Sciences partners
IBM Strategic Outsourcing
IBM HC and LS Development
 Teams of IT experts and SMEs
 Integrating standards
 Data Discovery and Query Builder
IBM Hardware, Software, Middleware
 Servers, Storage, Tivoli, Websphere, etc.
Page 53
IBM Research
 Med II
 IMR – integrated medical records
 GMS – Genomic messaging system
 Prima – patient record intelligent monitoring
and analysis
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Thank You
Q/A
Discussion
Page 54
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
BACK UP SLIDES
Page 55
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Clinical Genomics: Data Flow
Universities
Medical
Centers
Hospitals
CRO’s
Identifying
disease
markers
Proteomic
technologies
Sequence
detection
mechanisms
Transcrip.
profiling
Long-Term
Page 56
Population
sets (data
and samples)
Validating
disease
markers
Clinical
Research
(NDA’s and
diag.)
Short-Term
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Clinical Genomics: The Good News…Early Results
There are early and encouraging signs of new drugs and diagnostic
tests coming from a deeper understanding of genomics –
The first approved drug to come from an understanding of gene expression
was Gleevec from Novartis. It is used to treat chronic myeloid leukemia (about
20% of all adult leukemia) and had sales of $615M in 2002.
Another example is Genentech’s Herceptin treatment for breast cancer which is
an early example of personalized medicine
 Herception only works in women who have multiple copies of the Her-2/neu gene; approx. 25% of
all breast cancer patients. Sales in 2002 were $285M.
Other examples of how a deeper understanding of genomics can assist in
diagnostics and therapeutics
 Cerezyme, a drug for Gaucher disease found in Ashkenazi Jews with a specific genetic mutation.
A DNA test can identify this and indicate preventative treatment.
 Lotoronex, a drug for irritable bowel syndrome which was pulled from the market because of side
effect issues. GSK is looking at a diagnostic test which would predict those patients susceptible
to the side effects…otehrs could be treated successfully.
Source: Life Science Insights, 2004
Page 57
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Why IBM and Affymetrix?
Characteristics of an Information
Technology Partner
 Integrated solution provider with
sophisticated hardware, software and
services capability
 Leadership in middleware technology to
help set and propagate standards
 Extensive partner ecosystem in both
Healthcare and Pharmaceutical research,
and compatibility with their applications
 Vision and solution for information
technology challenges for entire
personalized medicine pipeline
 Domain expertise in data integration and
management and regulatory compliance
(HIPAA, 21 CFR Part 11, etc.)
 Scalable, open standards based solutions
that leverage existing investments in
research and clinical IT
Page 58
Characteristics of a Genomics
Technology Partner
 Platform available today and extensible to
the future
 Support for expression, genotyping and
resequencing
 Supports cost effective high throughput
applications
 Scalable from whole genome analysis to
focused clinical setting
 Standard for the community
 Regulatory “ready”
 Path to commercial diagnostics
 Proven leadership position in platform
technologies
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
Information Based Medicine: The Building
Blocks
The common thread for these pieces is the necessity of data
acquisition, management and analysis to improve diagnostic
decisions and patient outcomes.
Page 59
© Copyright IBM Corporation 2004
IBM Healthcare & Life Sciences
The Future of Healthcare lies in Prevention and
Personalization
Hospital
BioBank
Secure database
Genotype recorded
Baby
Physician’s office
Patient scenario
in the future
• Improve
quality of life
• Increase life
span
Personalized immunization
+
Screening schedule
+
Page 60
profiling
Preventive measures
Refined
treatment
Sources: Science, McKinsey
Screening
based on
biomolecular
Drugs
Life
Adult
Developed to target particular
disease subtypes on particular
genetic background
© Copyright IBM Corporation 2004