Transcript Slide - Maine Dartmouth Family Medicine Residency

Genomics, cancer, and medicine:
10 years after the
Human Genome Project
Edison Liu, M.D.
April 13, 2012
Declaration
We are entering one of the most
profound periods of advancement in
biology and medicine – one that will
transform:
Health and Medicine
Driven by technologies in genetics,
genomics, and computational biology
Genomics and Genetics
Genetics: study of genes and their function
Genomics: study of all genes and how they
function together
“Discover all possible genes involved in a
biological process or a human disease”
Genomics enables the
complete genetic “view” of a disease
Transformative technologies enabling this
revolution:
•New generation of ultra-fast sequencing
technologies
“Massive Data Generation”
• Enabling computational advances
“”Analytical power“
•Genetic engineering to model organisms of disease
“Surrogates of disease”
A*
C
T
A
G
C
T
G
A
T
T
C
A
G
T*
T*
T*
A
A
G
T
C
T
A
G
C
T
G
A
T
T A
C G
A T
G
G*
C*
A*
Massively Parallel
Sequencing:
Sequencing by
Synthesis
~105-107 kb per run = 0.1-3 X human genome equivalents / run
7 orders of magnitude
increase in throughput
MR Stratton et al. Nature 458, 719-724 (2009)
Cost of sequencing a human genome
300 million USD -
30 million USD -
3 million USD -
300 million
Cost-effective sequencing =
Accessibility of genomic data
1 million
300,000 USD $60,000
30,000 USD -
$ 3,000
Low complexity
image
Data complexity
when ordered
appropriately gives
high resolution
picture
Genomic scale information provides
unique biological insights
Genomics & Personalized Medicine
Towards personalized cancer medicine
No benefit
Non-selective
Conventional chemothrapy
Molecular marker
&
target
Severe side effect
Responders
Personalized cancer medicine
Other therapy
No treatment
Selection
High response rate
by marker
Molecular specific
drug
CAT ATE RAT
BAT ATE RAT
RAT ATE CAT
Wild Type
Point Mutation
Rearrangement
Cathryn
Kathryn
Kathrine
Polymorphism
Polymorphism
Polymorphism
Chronic Myelogenous Leukemia:
and the 9;22 chromosomal rearrangement
Molecular Mechanisms:
How to make a Leukemia
• Ph1 chromosome identified 1960 as a
marker for CML (Nowell)
• bcr-abl cloned and shown to be the
molecular mechanism 1984-1990
(Groffen and Lugo)
BCR
ABL
Chronic
Myelogenous
Leukemia
Chronic Myelogenous Leukemia:
Mortality 1969 - 2002
Large scale clinical
trials begin with Gleevec
Gleevec Approved
By FDA
STI521:Gleevec
BCR
ABL
From Concept to Molecular Mechanism
to Treatment
• Ph1 chromosome identified 1960 as a
marker for CML (Nowell)
• bcr-abl cloned and shown to be the
molecular mechanism 1984-1990
(Groffen and Lugo)
• Specific drug (Gleevec) to target gene
abnormality 1999 (Druker)
From discovery of a single oncogene
to treatment: 39 years
Nature 2007
In 2007, the genomic analysis of one lung
cancer from a 62 year-old smoker
 EML-ALK fusion in 6% of lung cancer patients
EML
ALK
Lung
Cancer
Crizotinib  60% response rate in those 6% of patients
with lung cancer with the EML-ALK mutation.
On August 26, 2011, the US FDA gave approval of
crizotinib by for the treatment of ALK-rearranged lung
cancer
4 years from genomic discovery to treatment
Ou, Drug Des Devel Ther. 2011; 5: 471–485.
Chronic Myelogenous Leukemia (CML)
Optimizing treatment for CML based on genetic
makeup of the patient
Clinical Challenge: Drug resistance
•Acquired resistance – resistance after long term treatment
- due to second ABL mutation
•Primary resistance – resistance at the beginning of
treatment.
•In Asia, complete cytogenetic response rates are lower 50% vs. 74%. Mechanism unknown
Question: is there a reason why 25% of CML cases do not
respond to imatinib?
Approach: We compared the genomes of three CML
cases with primary resistance to Imatinib with two
CML cases sensitive to Imatinib therapy
Results:
3/3 resistance cases
had the same 2.9kb
deletion in the BIM
gene not seen
in sensitive cases (0/2)
BIM:
• BIM is a gene that activates cell death (pro-apoptotic).
• Activated BCR-ABL1, suppresses BIM function thus
allowing leukemia cells to survive. When CML cells are
treated with Imatinib, BIM expression goes up  cell
death
Bcr-ABL: CML
Imatinib
Intact BIM
Death of
Leukemia cells
BIM deletion polymorphism:
• This deletion polymorphism is 3-5X more common in
CML cases resistant to imatinib that sensitive cases
• This 2.9 kb 2 deletion of BIM is not a mutation, but is a
polymorphism present in normal genomes (a germline
polymorphism):
12% in Asian individuals
0% in Africans
0% in Caucasians
How does it work?: The 2.9kb BIM deletion polymorphism results
an abnormal transcript (E3) that a produces a truncated and inactive
BIM protein
Normal
Transcripts
E3
Imatinib
Bcr-ABL: CML
Intact BIM
Death of
Leukemia cells
Imatinib
Bcr-ABL: CML
BIM
E3
Primary
Drug Resistance
We used this genomic intelligence to overcome
this resistance:
Imatinib
Bcr-ABL: CML
BIM 3
BH3 mimetics
Bcr-ABL: CML
Primary
Drug
Resistance
Imatinib
BIM
Death of
Leukemia
cells
This genomic experiment with 5 patients explains the lower response rate
In North Asians to a life saving treatment in CML.
Personalizing medicine in Asia
Now: New
~50% cytogenetic response
CML Patient
in Asia
Check for bcr-ABL
rearrangement
YES
Check for 2.9kb
deletion
polymorphism in BIM
YES
NO
Imatinib
&
BH3-mimetic
Imatinib
>75% cytogenetic
response
75% cytogenetic
response
Visualizing the Cancer Genome
Copy
LOH Number
Structural
Variations
BT55
‘Conductor’ mutation = early event that conducts the
direction of further cancer mutations
Mutation pattern
appears to be
generated by
separate cuts
when mapped to
the original
physical “map”
Chromosomal “origami” simultaneously
generates oncogenic “pattern”
Chromosomal origami to generate
cancer gene cassette
ERBB2
ERBB2
P53
ERBB2
17q21.3
ERBB2
17q21.3
ERBB2
17q21.3
ERBB2
BRCA1
Oncogenes
17q21.3
ERBB2/HER2
17q21.3 amplicon
Oncogene
BRCA1
Tumor
Suppressor
Gene
TD207
U-Inv331
U-Inv75
17q21.3 ERBB2
TD49 + Del67
Del51
There are 16 weak
oncogenes here.
4 that are synergistic
with ERBB2
oncogenesis
17q21.3
Amplification
BT55 (ER+, ERBB2++)
Luminal B
Chr17 ‘evolutionary origami’ has treatment implications for
Combination therapy
Cancer progression 
17p (TP53) loss
Chromosomal instability
Tandem
duplication in
ERBB2 locus
Recurrent UnpairedInversion:
“Conductor”
Mutation
Massive ERBB2
amplification
Nutlins
17q21.3
amplification
BRCA1 locus loss
Oncogenes
Lapatinib
Tumor
Suppressor
genes
New Target
PARP inhibitor
Cancer Genomics Consultation Model Using
Mouse Avatars for Human Disease
Druggable
mutations
Tumor DNA
sample
Sequence Tumor
Germline DNA
sample
Generation of
serum-based
personalized and
private cancer
biomarker test
Expand tumor in
NSG mice
Automated
sequence
analysis of tumor
and germline
Germline
pharmacogenetic
analysis
Identification of
cancer specific
rearrangements
Monitor for
recurrence and
clonal variation
Test specified
drugs for
response in vivo
Prognostic
information
Devise optimal
combination
therapy
Visualization
formatted report
Consultation
with patient and
physician
for treatment
plan
Radiologist
of the Genome
Radiologist
Interprets complex data
rendered through
computational algorithms
Is the consultants
to doctors
Genomics, cancer, and medicine:
10 years after the
Human Genome Project
Edison Liu, M.D.
April 13, 2012
Age Adjusted Mortality for
breast cancer is declining since 1990
D ~ 20%
1990
T2N1M0
Age: 56 family history: negative
Sequence performance:
ER (IHC) positive
HER2 (FISH) Positive
Ki67: +++
GHI recurrence score: XXX
9.6 million reads; 75 base pair, paired end on Illumina HiSeq
Comparison of germline and cancer genomes
This is a visual representation of the cancer genome of your
patient as compared to her constitutional (germline) genome.
The 23 chromosomes are arrayed in a circle; amplifications
are on the outer circle, and the deletions and inversions are
in the inner circles. In the innermost circle is a representation
of the chromosomal translocations. Each arc represents a
translocation, and the intensity of the arcs is an indication of
Amplification of that translocation. By clicking onto the figure,
you will get a blow up of the schematic and a detailed legend.
Maximal
Normal
The mutational load score is a composite score the integrates
the mutational load that is seen in the tumor of your patient.
It is made up of two components: sequence mutations, structural
mutations/rearrangements. Your patient’s mutational score when
compared to a panel of XXX tumors of the same type is in the
following distribution:
What if this trend continues?
1990
Can death from
breast cancer be
eliminated?
2020