Cancer genomics & precision medicine in the 21st century

Cancer genomics & precision medicine in the 21st century

Loading
Loading Social Plug-ins...
Language: English
Save to myLibrary Download PDF
Go to Page # Page of 51

Description: Single Nucleotide Polymorphisms (SNP): Variation in single base in DNA in germline, most common variants in genome (over 50 million identified), SNP arrays interrogate the entire genome -uses DNA from germ-line (blood). Used in Genome Wide Association Studies (GWAS): Typically uses SNP arrays to compare populations (with disease or not), Determines risk or susceptibility to some state. Cancer is a disease of the genome (cont): Driver mutations are the mutations we would like to target and inhibit their function.

 
Author: Lee J. Helman MD (Senior) | Visits: 539 | Page Views: 842
Domain:  Medicine Category: Therapy 
Upload Date:
Link Back:
Short URL: https://www.wesrch.com/medical/pdfME1LYY000AZJS
Loading
Loading...



px *        px *

* Default width and height in pixels. Change it to your required dimensions.

 
Contents:
Cancer Genomics & Precision Medicine
in the 21st Century
Lee J. Helman, MD Scientific Director for Clinical Research
CCR, NCI

Outline
• Define terms
• Describe vision for how genetic
characterization of tumors will change
treatment paradigms for cancer in the
future
• Describe ongoing clinical trials currently
using this approach
• Describe potential difficulties/pitfalls

Terms/Vocabulary
• Single Nucleotide Polymorphisms (SNP)
• Variation in single base in DNA in germline,
most common variants in genome (over 50
million identified)
• SNP arrays interrogate the entire genome-uses
DNA from germ-line (blood)

• Used in Genome Wide Association Studies
(GWAS)
• Typically uses SNP arrays to compare
populations (with disease or not)
• Determines risk or susceptibility to some state

Terms (con’t)
• RNA expression profiles-determines global
messenger RNA expression in a sample-using
hybridization of mRNA to a Chip
• Methylation arrarys-determines global
methylation of the genome-an epigenetic
change typically inserts a methyl group at
CpG islands in DNA and alters transcriptionusing hybridization of DNA to a Chip
• Massively parallel sequencing-allows for rapid
sequencing of entire exome (WES) on whole
genome (WGS) or cDNA (RNA-seq)

Possible Master
Cancer Susceptibility
Region 8q24

127.6 M

GWAS

rs979200

Region 2
Prostate
only

rs1456310
rs6993569
rs6470494

Breast only
Region 3
Prostate & Colon
Region 1
Prostate only

rs13281615
rs16902124
rs16902126
rs6983267
rs10505476
rs7837328
MYC

rs1447295
rs7837688

129.0 M

Genetic Predisposition to Breast
Cancer European Population
Population genotype relative risk

10

3

BRCA1
BRCA2
TP53
PTEN
CHEK2
ATM
PALB2
BRIP1

Uncommon
Rare

1.5
1.4
1.3
1.2
1.1
1
0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

5p12
5q11.2

16q12.1

8q24

10q26

Population risk-allele frequency

2q35

1

New Loci
CGEMS
CGEMS/BCAC

Expression profile Identification
of Breast Tumor Intrinsic Subtypes

Carey, L. A. et al. JAMA 2006;295:2492-2502.

Copyright restrictions may apply.

SDH Deficient GIST Have Global
Hypermethalyation

Massively Parallel Sequencing
(Next-Generation Sequencing)
Genomic DNA
or RNA

Fragmentation

Fragment

Size Selection

DNA Fragments of
Similar Sizes

Adaptors Ligation
Genomic DNA Library

Amplification and Sequencing
Ref. Genome
AGCTGCTCGTCGCGAAACTCCGATCGACTGCTGATCGACTCGATCACTCGATCGTAGTCGAGAGTACTCGATGCT

Align (Map) Reads
to Ref. Genome
Genome Sequence

Types of Alterations that can be Detected using

Next Generation Sequencing will Identify Other Driver
Mutations and Enable Individualized Therapy for Cancer
Therapy

Roche / 454
Genome Sequencer FLX
Titanium

Life Technologies Life
SOLiD v4
Technologies
5500 XL

Illumina / GAll/HQ
2500
Whole Genome 48
hrs

Life
Technologies
Ion Torrent

Life
Technologies
Proton
1 Genome 2
hrs

PacBio RS
Ion Torrent

Helicos
HeliScope

Comprehensive Analysis of the
Rhabdomyosarcoma genome: Study Design
Discovery set
Whole Genome Sequencing (WGS)
(Complete Genomics Platform)
46 RMS
(22 ARMS)

SNP Array
HumanOmni2.5-8 BeadChip
134 RMS (38 ARMS)
30 Overlap with CG WGS

Validation set
Whole Exome (Agilent, SOLiD,
Illumina)
133 RMS (52 ARMS)
30 Overlap with CG WGS

2/46 Samples had Aberrant
Fusions
Incorrect Diagnosis
1. Sample 1 –Original Histology: Sarcoma, Un
differentiated RMS, Not Otherwise Specified. Had
ALK-NPM1 fusion by whole genome sequencing;
Review of Histology = Hematological malignancymost likely misdiagnosed ALCL
2. Sample 2 - Original Histology: Consistent with
RMS Presence of RET-NCO4 fusion by whole
genome sequencing which has been reported in
papillary thyroid carcinoma. Likely misdiagnosed
or sample label error from source

Typical Fusion Positive ARMS; Fusion Gene,
Low Aneuploidy, Low Somatic Mutation rate
t(2;13)

11pLOH

Typical ERMS; Higher Aneuploidy, Higher
Somatic Mutation rate

ARMS-Fusion Gene Detection:

• 22-ARMS by
Histology
• 14-PAX3-FOXO1
• 3-PAX7-FOXO1
• 1-PAX3-NCOA1
• 1-Novel
• 3-Fusion Negative

Fusion Negative ARMS Shows massive 2q
Rearrangement with in-Frame PAX3-INO80D

Purple:tail-tohead Green:
head-to-tail
junction (possibly
tandem
duplication)
Orange: tail-to-tail
junction or headto-head junction
(inversion)

RNAseq Confirms Expression of PAX3INO80D
Novel Fusion Transcript

HOW HAS THIS CHANGED
CANCER TREATMENT?

Change in View of Lung Cancer

Shifting the Paradigm
Previous Approach

New Practices

Descriptive medicine

Understanding of disease
mechanisms

Empirical diagnosis

Mechanism-based
diagnosis/treatment

Grouped by Organ Site

Sub-grouped by
molecular/biological
classification

Uniform treatment

Individualized treatment

Retrospectively diagnose
disease

Prospectively evaluate
relative disease risk

Acute care

Early detection and
intervention

Toward Precision Medicine
Put more science into clinical trials
Investigational drug
Responders
Non-responders

Pharmaco dynamic
measurements

Molecular diagnostics:
candidate approach

Modified from American Association for Cancer Research

Molecular diagnostics:
unbiased approach

Precision Medicine
Breast cancer
Treatment A
Molecular
diagnostics
Prostate cancer

Treatment B

Treatment C
Lung cancer

Standard TX
Modified from American Association for Cancer Research

Cancer is a disease of the
genome
• Therefore, if we precisely define the cancer
genome, we will understand and cure cancer
• Why we must be cautious about such statements

• Definitions
• Founder mutations-first genomic mutation
• These are often lesions that lead to
genomic/chromosomal instability (p53, RB, etc.) and
are often not fully transforming

• Driver mutations-these are mutations that are
required for expression of fully transformed
phenotype

Cancer is a disease of the genome
(cont)
• Driver mutations are the mutations we would like to
target and inhibit their function

• Passenger mutations-these mutations are
“collateral damage” resulting from genomic
instability and are not required for
maintaining the transformed phenotype,
therefore are “noise” in the system
• Since most cancers are rapidly evolving
biologic entities, it is a major task to sort out
“drivers” from “passengers”, and these may
change over time

Mutation Rates across Cancers are not Uniform
100 / Mb

n=53

7

55

36

200

3

45

82

9

28

21

40

20

10 / Mb

1 / Mb

RMS

0.1 / Mb

??

SMOKING

UV

Courtesy of Gad Getz

Clonal Evolution
Founder RB1 and p53 mutations followed by additional mutations

Greenman C D et al. Genome Res. 2012;22:346-361

©2012 by Cold Spring Harbor Laboratory Press

Signaling pathways are not 1-way

• Driver mutations in signaling pathways
(kinases) are components of highly integrated
“wiring” that is not a one way flow of
information
• Because these are critically important for normal cell
functions, these are highly regulated pathways

• Perturbation of a single component of will
lead to activation of other components due to
feedback activation or loss of feedback
repression

Kinase oncogene dependence and principles of drug resistance.

Wagle N et al. JCO 2011;29:3085-3096

©2011 by American Society of Clinical Oncology

Example of Vemurafenib

• 50-60% of melanoma patients have driver
mutations in BRAF (V600E)
• At doses of vemurafenib that inhibit 90% of BRAF activity, most patients respond rapidly
with tumor shrinkage
• Median duration of response is less than 12
months due to resistance
• What are the mechanisms of resistance?

Example-BRAF (V600E) mutations in
colon cancer

• Unresponsiveness of colon cancer to BRAF
(V600E) inhibtion through feedback activation
of EGFR Prahallad A, et al. Nature Jan 26 2012
• Mechanism-appears to be inhibition of BRAF
leads to inhibition of MEK and ERK, leading to
reduced phosphatase activity of CDC25C,
leading to reduced dephosphorylation of
EGFR, leading to increased activation and
EGFR signaling

Neuroblastoma
• One of the SRBCT
• Derived from primordial neural crest cells destined to
become sympathetic ganglia in the peripheral nervous
system not CNS
Incidence:
• 1 per 100,000 in children < 15 yrs in US (650 cases per year)
• The most common extra cranial solid tumor for children
• 7-10% of cancers of childhood
Survival Rates:
• 95-70% for low stage tumors (1,2,3)
• ~50% patients present with advanced disease
• < 30% of children over 1 year old with advanced disease
and/or MYCN amplification despite aggressive therapy

Patient (19yr) with High-risk Neuroblastoma
Diagnosis
~4 Months

4
cycles
Inducti
on
 A3973

Surgery
3 years

 8 cycles of
Salvage
Therapy
 7 cycles of
RA
 Radiation
multiple
sites
 Low dose
MIBG

Death

Bone Marrow

• 19 yrs old
• Stage 4

Met1-BM:
bone marrow
biopsy at
diagnosis.
>80% tumor

Primary: tumor
removed by
surgery
viable margin

Met2-Liver:
autopsy
Whole genome seq of liver Met2
macro-dissected
& RNAseq of Met1, primary and
Met2

Forty-four (44) non-synonymous mutations
found in index sample (Liver Met)

Chromothripsis was evident by massive
complex rearrangements detected at
chromosomes 4q and 13p

Ion Torrent: Deep Re-sequencing (1000x) of primary
(bone marrow) and 4 primaries (adrenal): 14/44 (32%)
small variants were present in all samples, 30 unique to
liver met

RNAseq of Met1, Primary, Liver Met2 to
identify expressed driver mutations:
• 14/44 commonly mutated in all 3 tumors
• 12 the gene is expressed
• 9 variant allele expressed
• 3 genes (NUFIP1, GATA2, and LPAR1) high variant allele fraction
>30%

30/44 Somatic Mutations Unique to Liver Met2
• 11/30 the variant was expressed in Liver Met2
• De-Novo mutations in liver arising during the course of disease but
absent in primary (4 regions) and bone marrow met

Summary
1. Neuroblastoma is marked by aneuploidy
in recurrent regions but lack frequently
recurring mutations
2. Possible that classic mutations may not
drive tumorigenesis
3. Possible that each individual tumor has its
own set of driver mutations
4. Ongoing efforts including RNAseq
underway and will identify key onocogenic
drivers and targets for therapy

Many Novel Drivers Epigenetic (red)
Many Not Currently Druggable
PBRM1 –
Renal cell carcinomas
EZH2, MEF2B –
Lymphomas
KCNJ5 –
Adrenal adenomas
DNMT3A, SF3B1, SRSF2, U2AF35 – Leukemias
MLL2, MLL3, DDX3X – Medulloblastomas
ARID1A, PPP2R1A – Ovarian cancers
DAXX, ATRX – Pancreatic endocrine tumors
BAP1, TTRAP, PREX2 –Melanomas
IDH1, 2 –
Gliomas
CIC, FUBP1 – Oligodendrogliomas
MED12 ‐ Leiomyomas
H3F3A, HIST1H3B- Diffuse intrinsic pontine glioma
ATRX, ARID1A, ARID1B, PTPN11- Neuroblastoma

Adapted from Vogelste

M-PACT: Molecular Profiling based
Assignment of Cancer Therapeutics
Pilot Trial to Assess the Utility of Genetic
Sequencing to Determine Therapy and
Improve Patient Outcome in Early Phase
Trials
NCI-Sponsored Clinical Trial

Objective
• Assess whether the response rate (CR+PR) and/or 4month PFS is improved following treatment with agents
chosen based on the presence of specific mutations in
patient tumors.
– Only patients with pre-defined mutations of interest will be
eligible
– Study treatments, regardless of cohort, will be chosen from the
list of regimens defined in the protocol
– Arm A: Receive treatment based on an study agent prospectively
identified to work on that mutation/pathway
– Arm B: Receive treatment with one of the study agents in the
complementary set (identified to not work on one of the
detected mutations/pathways)
46

Patient Population





Patients with refractory solid tumors that have progressed on at least one
line of standard therapy or for which no standard treatment is available
that has been shown to improve survival.
Adequate organ function ( AST/ALT 100K, ANC> 1500)
Study regimens: As long as the same set of protocols are offered to a given
set of patients, the number and actual treatments regimens can vary over
time
Mutations in DNA repair pathways

Veliparib+ Temozolomide
MK1775 + carboplatin

Mutations in the PI3K pathway; loss of
PTEN, Akt amplification

mTOR inhibitor -Everolimus

Mutations in the RAS pathway

GSK 1120212 (MEK inhibitor)
47

Study Design

Biopsy
Mutation
detected
Sequence
fresh biopsy
tissue from
all patients

Mutation
not
detected

Off-Study

Randomly
assign pt to Arm
A or B if
actionable
mutation
identified
(Clinical team
blinded to the
specific
mutation data)

Arm A:
Targeted
therapy based
on the
patient’s
mutational
data

Arm B: Therapy
not
corresponding
to the patient’s
mutational
data

Assign
protocol

Assign
protocol
(allow cross
over at
progression to
the targeted
agent)

Statistical Design
• Patients will be randomized 2:1 to Arm A
(experimental) versus Arm B (control)
• Within Arm A, up to 30 patients will be treated
within each of the treatment cohorts. Within each
treatment cohort of Arm A, discriminate between
tumor response rates of 20% vs. 5% and, as a
secondary endpoint, 4-month PFS rates of 50% vs.
25%
• The two Arms will be compared with respect to both
objective response rate and PFS – this is a
randomized comparison.
49

Gene Name
BRAF
NF1
Kras

Pathways/Function

Gain or Loss of Function?

Genes and Pathways of Interest

Hras

RAS/RAF/ERK/MEK
RAS
RAS/RAF/ERK/MEK
AKT/PI3K
RAS/RAF/ERK/MEK
AKT/PI3K
RAS/RAF/ERK/MEKAKT/PI3K

Gain

AKT1
AKT2
AKT3
PIK3CA

AKT/PI3K
AKT/PI3K
AKT/PI3K
AKT/PI3K/RAS/RAF/ERK/MEK

Gain
Gain
Gain
Gain

PTEN

AKT/PI3K/RAS/RAF/ERK/MEK

Loss

P53
FBXW7
ATM
PARP1
PARP2
ERCC1
MLH1
MSH2
NBN
ATR
MGMT

DNA Repair
DNA Repair
DNA Repair
DNA repair
DNA repair
DNA repair
DNA repair
DNA repair
DNA Repair
DNA repair
DNA repair

Loss
Loss
Loss
Loss
Loss
Loss
Loss
Loss
Loss
Loss
Loss

Nras

Gain
Loss
Gain
Gain

Conclusions
• The ability to obtain full genomic data on a
given tumor will allow us to make rational
choices for therapy
• Functional genomics may provide help in
choosing combination therapy
• Combinations will not be easy due to enhanced
toxicities

• Cancer as a chronic disease is not a bad thing
as long as we recognize rapid development of
resistance and clonal evolution