Clinical Cancer GenotypingClinical Cancer Genotyping

Long Phi Le, MD/PhDLong Phi Le, MD/PhD

Department of PathologyDepartment of Pathology

Diagnostic Molecular Pathology LaboratoryDiagnostic Molecular Pathology Laboratory

Translational Research LaboratoryTranslational Research Laboratory

Massachusetts General HospitalMassachusetts General Hospital

Boston, MABoston, MA

lple@partners.orglple@partners.org

The case for broad cancer genotypingThe case for broad cancer genotyping

Our current approachOur current approach

Next generation sequencingNext generation sequencing

SOLiDSOLiD / Ion Torrent Pilot/ Ion Torrent Pilot

The Old Paradigm in Cancer TreatmentThe Old Paradigm in Cancer Treatment

The New Paradigm in Cancer TreatmentThe New Paradigm in Cancer Treatment

t(9;22) in CMLt(9;22) in CML

Blast crisis

Obrien et al., Imatinib Compared

with Interferon and Low-Dose

Cytarabine for Newly Diagnosed

Chronic-Phase Chronic Myeloid

Leukemia. N Engl J Med 2003

-100

0

100

200

300

400

500

600

700

800

900

Well

Cycle

4

Cycle

8

Cycle

12

Cycle

16

Cycle

20

Cycle

24

Cycle

28

Cycle

32

Cycle

36

Cycle

40

Cycle

44

FISH detects HER2 amplification in

15-25% of ductal carcinomas

Predicts responsiveness to

Trastuzumab

HER2 Amplification in Breast CancerHER2 Amplification in Breast Cancer

Romond EH et al., Trastuzumab

plus Adjuvant Chemotherapy for

Operable HER2-Positive Breast

Cancer. N Engl J Med 353:1673.

2005.

EGFR: Erlotinib/ Gefitinib

20% Lung adenocarcinomas

KRAS: Cetuximab resistance

36-50% Colon adenocarcinoma

Molecular Markers and Targeted TherapyMolecular Markers and Targeted Therapy

ALK: Crizotinib

3-5% Lung adenocarcinoma

BRAF V600E: PLX4032

60% Melanoma

BRAF

1799 T>A

V600E

DRUG MECHANISM OF ACTION ACTIVE MGH STUDIES

Axitinib PDGFR, VEGFR Inhibitor 1

AZD-2171/Cediranib VEGFR Inhibitor 4

AZD-6244 MEK Inhbitor 1

Bevacizumab Anti-VEGF 15

BI-2536 Plk-1 Inhibitor 1

Cetuximab Anti-EGFR 6

Dasatinib Bcr-Abl, Src Kinase Inhibitor 2

Erlotinib hydrochloride EGFR Inhibitor 6

Enzastaurin hydrochloride PKC beta inhibitor 1

Gefitinib EGFR Inhibitor 4

Lapatinib ditosylate EGFR, HER2 Inhibitor 1

Targeted Molecular Therapy in Active MGH TrialsTargeted Molecular Therapy in Active MGH Trials

Neratinib EGFR, HER2 Inhibitor 1

PF-02341066 MET Inhibitor 1

Rituximab Anti-CD20 3

Sirolimus mTOR Inhibitor 1

Temsirolimus mTOR Inhibitor 1

Trastuzumab Anti-HER2 1

CP-751871 Anti-IGF-1R 1

EXEL-7647 EGFR, HER2, EphB4, VEGFR Inhibitor 2

EXEL-2880 MET, VEGFR-2 Inhibitor 1

AZD-6244 MEK Inhibitor 1

Everolimus mTOR, FKBP12 Inhibitor 1

Tandutinib PDGFR, KIT, Flt3 Inhibitor 1

Sorafenib Multi-kinase Inhibitor 2

Sunitinib Flt3, VEGFR, KIT, PDGFR Inhibitor 11

Vatalanib VEGFR-2, KIT, PDGFR Inhibitor 2

Volociximab Integrin Inhibitor 1

Metastatic Colorectal Cancer & KRAS TestingMetastatic Colorectal Cancer & KRAS Testing

“Based on systematic reviews of the relevant literature, all patients

with metastatic colorectal carcinoma who are candidates for anti-

EGFR antibody therapy should have their tumor tested for KRAS

mutations in a CLIA-accredited laboratory. If KRAS mutation in

codon 12 or 13 is detected, then patients with metastatic colorectal

carcinoma should not receive anti-EGFR antibody therapy as part of carcinoma should not receive anti-EGFR antibody therapy as part of

their treatment.”

-Allegra et al. J Clin Oncol 2009

US Growth in Genetic TestingUS Growth in Genetic Testing

MolecularMolecular

CytogeneticsCytogeneticsMicroarraysMicroarrays

Molecular Diagnostic TechniquesMolecular Diagnostic Techniques

PCR and QPCRPCR and QPCR SequencingSequencing

Genomic

TL-09-267 20 ng/panel DNA

More With LessMore With Less

TL-09-267 20 ng/panel DNA

TL-09-285 3.04ng/panel DNA

Considerations in Clinical Genotyping PlatformConsiderations in Clinical Genotyping Platform

Clinical Test

• Archived FFPE tissue

• Analytical sensitivity

• 2 week turnaround time

• Report in patient’s record

• Performed in a CLIA lab

Actionable Targets

• Predicts response/resistance

• Clinches diagnosis

• Yields prognosis

• Stratify patients for trials

• Adaptability for new targets

Logistics

• Clinical patient coordinator

• Accessioning

• Automation

• Eventually test all tumors

Other

• Economics

• Finances and billing

• Bioinformatics

• Potential for research

Distribution of KRAS MutationsDistribution of KRAS Mutations

COSMIC, Wellcome Trust Sanger Institute, 2010

ddNTP

ddNTP

ddNTP

loci of interest

Multiplex PCR Single Base Extension Reaction Capillary Electrophoresis

SNaPshot GenotypingSNaPshot Genotyping

Electrophoretic Output

Increasing

molecular weight

Re

lative

flu

ore

sce

nce

A B DC FE

Gene Amino Acid – cDNA Residue Gene Amino Acid – cDNA Residue AKT1 49G - E17 KRAS G12 - 34G KRAS G12 - 35G APC R1114 - 3340C KRAS G13 - 37G APC Q1338 - 4012C KRAS G13 - 38G APC R1450 - 4348C APC T1556fs* - 4666_4667insA NOTCH1 L1575 - 4724T NOTCH1 L1601 - 4802T BRAF V600 - 1798G BRAF V600 - 1799T NRAS G12 - 34G NRAS G12 - 35G CTNNB1 D32 - 94G NRAS G13 - 37G CTNNB1 D32 - 95A NRAS G13 - 38G CTNNB1 S33 - 98C NRAS Q61 - 181C CTNNB1 G34 - 101G NRAS Q61 - 182A CTNNB1 S37 - 109T NRAS Q61 - 183A

AKT1 49G – E17

MGH SNaPshot AssayMGH SNaPshot Assay

CTNNB1 S37 - 109T NRAS Q61 - 183A CTNNB1 S37 - 110C CTNNB1 T41 - 121A PIK3CA R88 - 263G CTNNB1 T41 - 122C PIK3CA E542 - 1624G CTNNB1 S45 - 133T PIK3CA E545 - 1633G CTNNB1 S45 - 134C PIK3CA Q546 - 1636C PIK3CA Q546 - 1637A EGFR G719 - 2155G PIK3CA H1047 - 3139C EGFR T790 - 2369C PIK3CA H1047 - 3140A EGFR L858 - 2573T PIK3CA G1049 - 3145G EGFR E746_A750 - 2235_2249del EGFR E746_A750 - 2236_2250del PTEN R130 - 388C EGFR Exon 19 deletions PTEN R173 - 517C PTEN R233 - 697C FLT3 D835 - 2503G PTEN K267fs*- 800delA IDH1 R132 - 394C TP53 R175 - 524G IDH1 R132 - 395G TP53 G245 - 733G TP53 R248 - 742C JAK2 V617 - 1849G TP53 R248 - 743G TP53 R273 - 817C KIT D816 - 2447A TP53 R273 - 818G TP53 R306 - 916C

ERBB2 Exon 20 insertions

IDH1 R132 -394C

IDH1 R132 -395G

EGFR2235_49R

NRAS38

NRAS182

PI3K263 bCat95bCat122

TP53.742

Genomic DNA

BRAF1799

MGH SNaPshot v1.0MGH SNaPshot v1.0

Melanoma

BRAF

1799 T>A

V600E

0.44

TP53 5%

IDH1 <1%

NRAS 1%BRAF 2%

HER2 2%

PIK3CA 4%

ALK 3%

CTNNB1 2%

AKT 1%

Lung Cancers: SNaPshotLung Cancers: SNaPshot

KRAS 23%

No Mutation 42%

EGFR 15%

TP53 5%

N=650

KRAS

56 isolated

(58 total)

EGFR

36 isolated

(50 total)

PIK3CA

51 1

13

B-cat

Lung Cancers: Mutation OverlapLung Cancers: Mutation Overlap

1

2

(58 total)

ALK

13

T790M

5TP53 11

14

2

APC

NRAS

BRAF

13

No Mutation Identified34%

PIK3CA 6%

NRAS 3%

APC 4%

BRAF 7%

Colorectal Cancers: SNaPshotColorectal Cancers: SNaPshot

KRAS25%TP53

21%

N=250

PIK3CA

BRAF

6 isolated

4

3

1

1

3

1

Colorectal Cancers: Mutation OverlapColorectal Cancers: Mutation Overlap

KRAS

20 isolated

(36 total)

TP53

18 isolated

(28 total)

APC

1 NRAS

3

4

261

More Than Just Point MutationsMore Than Just Point Mutations

The Future of Clinical Cancer GenotypingThe Future of Clinical Cancer Genotyping

Do we have the technology?

Is it cost-effective?

What to genotype?

The challenges?

By Angela Canada Hopkins

Next Generation Sequencing: The Big BrothersNext Generation Sequencing: The Big Brothers

Roche 454 Illumina/Solexa

Life Technology SOLiD Helicos

Next Generation Sequencing: Uno Tube, Mucho DataNext Generation Sequencing: Uno Tube, Mucho Data

Next Generation Sequencing

First Generation Sequencing

Ten Years of SequencingTen Years of Sequencing

Adapted from Nature 1 April 2010

Next Generation Sequencing: The Little BrothersNext Generation Sequencing: The Little Brothers

Roche 454 GS Jr Illumina GA IIe

Life Technology SOLiD PI Life Technology Ion Torrent

The Future of Clinical Cancer GenotypingThe Future of Clinical Cancer Genotyping

Do we have the technology?

Is it cost-effective?

What to genotype?

The challenges?

By Angela Canada Hopkins

How Many Genes?How Many Genes?

Pre-Human Genome Project

• >100,000 genes

Post-Human Genome Project

• 30,000 genes• 30,000 genes

• 20,000-25,000

Human exome

• ~180,000 exons � 1% of the

genome � ~30Mb

Clinical Cancer Genotyping: A Mountain to Climb?Clinical Cancer Genotyping: A Mountain to Climb?

Published Cancer Exomes

• 11 Colorectal – Science 2007

Bert Vogelstein:

AACR 2010 Meeting

Plenary Session

Mu

tati

on

s p

er

Tum

or

Non-Silent Mutations in Pancreatic Cancer

• 11 Colorectal – Science 2007

• 11 Breast – Science 2007

• 24 Pancreas – Science 2008

• 22 Gliomas – Science 2008

• 2 Leukemias – NEJM, Nature 2008

• 1 Breast – Nature 2010

• 1 Breast – Nature 2009

• 4 Granulosa Cell – NEJM 2009

• 1 Lung – Nature 2010

• 1 Melanoma – Nature 2010

• 22 Medulloblastomas - Unpublished

Mu

tati

on

s p

er

Tum

or

Mu

tati

on

s p

er

Tum

or

Non-Silent Mutations in Different Tumors

Clinical Cancer Genotyping: Focus! Focus!Clinical Cancer Genotyping: Focus! Focus!

Review of Literature/Databases

• 116,432 human cancers

Bert Vogelstein:

AACR 2010 Meeting

Plenary Session

Mu

tati

on

s p

er

Tum

or

Genetic Alterations in Pancreatic Cancer

• 116,432 human cancers

• 353 histopathologic subtypes

• 130,072 intragenic somatic mutations

• 3142 mutated genes

Potential Driver Genes

• 286 tumor suppressor genes (>15% of

mutations are truncating)

• 33 oncogenes (same codon mutated in

at least 2 tumors)

Mu

tati

on

s p

er

Tum

or

Driver Gene Alterations in Pancreatic Cancer

A Little MathA Little Math

Desired Analytical Sensitivity

• 1-5%

Typical NGS Error Rate

• 1-2%

Whole Genome Sequencing

• 30x• 30x

• 1 error � >3.3% sensitivity

Targeted Sequencing

• 200-500x

• 0-4 errors in 200 reads � 1%-2%

error

• Set threshold at ≥5%

Whole Genome Sequencing at 200x

• >$60,000!

Target Exon CaptureTarget Exon Capture

Microdroplet PCR Solid-phase Capture Solution-phase Capture

Metzker ML, Nature Reviews Genetics 2010

Reaction Array

Cost of Next Generation SequencingCost of Next Generation Sequencing

Adapted from Nature 1 April 2010

Library Preparation Cost

The Future of Clinical Cancer GenotypingThe Future of Clinical Cancer Genotyping

Do we have the technology?

Is it cost-effective?

What to genotype?

The challenges?

By Angela Canada Hopkins

SOLiD Sequencing Pilot Study on SNaPshot SamplesSOLiD Sequencing Pilot Study on SNaPshot Samples

• Emulsion PCR used to template

each sample on to magnetic beads.

• Beads modified at 3’ end to allow

for deposition on to sequencing slide.

• PCR amplicons concatenated and

sheared for barcoded fragment library

construction.

•Samples pooled prior to emulsion PCR

Courtesy of Life Technologies

50 bp

• 5 mer barcode sequenced first to parse each library

• Sequence 50 bases sequenced in the forward direction.

• Mapping and SNP calling performed in CLC Bio Genomics

Workbench (ungapped local alignment in colorspace, SNP

cutoff at 5%)

• Deposited slide loaded into one of two

available flow cells on SOLiD instrument.

5 bp BC

SOLiD Sequencing Pilot ResultsSOLiD Sequencing Pilot ResultsC

ase

Tota

l

Re

ad

s

% Ma

pp

ed

Min

imu

m

Co

vera

ge

Ma

xim

um

Co

vera

ge

Ave

rag

e

Co

vera

ge

1

8,551,464 35.2 804 113000 28691

2

8,380,102 35.9 851 126000 29282

3

9,700,737 35.7 1270 123000 32229

SOLiD

Next Generation

Sequencing

Variant Calls

SNaPshot

Single Base Extension

Genotyping Results

KRAS c.34G>T (30.1%)

EGFR c.2235_2249del15

KRAS c.34G>T

EGFR 15 bp del

TP53 c.743G>T (26.0%)

TP53 c.743G>T

KRAS c.34G>A (16.4%)

TP53 c.536A>T (10.4%) KRAS c.34G>A9,700,737 35.7 1270 123000 32229

4

7,487,505 35.2 905 100000 24460

57,447,964 34.7 913 84008 24020

6

7,424,530 35.1 189 116000 25268

7

7,788,914 34.9 185 135000 26097

8

7,748,550 35.3 281 130000 25881

9

9,260,386 34.7 283 146000 30972

TP53 c.536A>T (10.4%) KRAS c.34G>A

NRAS c.182A>G (47.7%)

TP53 c.880G>T (63.3%)

EGFR c.2240_2257del18

NRAS c.182A>G

EGFR 18 bp del

KRAS c.34G>C (63.3%) KRAS c.34G>C

KRAS c.38G>A (22.6%)

PIK3CA c.1633G>A (18.8%)

TP53 c.818G>A (39.9%)

KRAS c.38G>A

PIK3CA c.1633G>A

TP53 c.818G>A

BRAF c.1799T>A (22.1%)

PIK3CA c.1636C>A (14.4%)

EGFR c.2264C>A (7.4%)

BRAF c.1799T>A

PIK3CA c.1636C>A

TP53 c.743G>A (2.3%)

TP53 c.752T>G (1.9%) TP53 c.743G>A

KRAS c.35G>T (12.0%)

TP53 c.713G>T (20.9%) KRAS c.35G>T

KRAS Exon 2: Per Base Error & CoverageKRAS Exon 2: Per Base Error & Coverage

Per Base CoveragePer Base Coverage

Average Per Base Error

+/- Standard Deviation

Average Per Base Error

+/- Standard Deviation

Per Base Error

95% Confidence Interval

Per Base Error

95% Confidence Interval

Circles = Outliers Outside

Per Base Error 95% CI

Circles = Outliers Outside

Per Base Error 95% CI

KRAS Exon 2: Per Base Error & CoverageKRAS Exon 2: Per Base Error & Coverage

KRAS Exon 2: Per Base Error & CoverageKRAS Exon 2: Per Base Error & Coverage

#1

#5

#4

#3

#2

#9

TP53 Exon 7: Detection of Compound HeterozygosityTP53 Exon 7: Detection of Compound Heterozygosity

#8

Detection of Compound HeterozygosityDetection of Compound Heterozygosity

Detection of Compound HeterozygosityDetection of Compound Heterozygosity

468 mutations

10 mutations

Sanger COSMIC Website

Sample #1 Sample #4

EGFR Deletion: MappingEGFR Deletion: Mapping

Digital Genotyping and Structured DataDigital Genotyping and Structured Data

CDS Variant aa Variant Confidence

Score

Allele

Frequency

Relevance Prevalence Literature

BRAF

c.1799G>T

BRAF V600E 95 45% (1) Oncogenic

(2) Response

to PLX4032

2% Lung ACA

7% Colon ACA

60% Melanoma

Flaherty et al.

NEJM 2010

Laboratory (genotype):

Variant Confidence Score = f(Quality Value, Coverage, Error, Bi-directionality)

Patient Diagnosis Stage Genotype Treatment Response Literature

Jane Doe Metastatic

melanoma

III BRAF

c.1799G>T

Plexxikon

PLX4032

81% Melanoma Flaherty et al.

NEJM 2010

Clinical (phenotype):

Clinical Cancer Genotyping: At The Tipping PointClinical Cancer Genotyping: At The Tipping Point

Clinical targeted sequencing of

FFPE DNA

• initially 100 exons � >1000

• 200-500X coverage

• 100-150+ Mb data

• 3-4 week turnaround time

• $500 raw reagent cost• $500 raw reagent cost

Desired

• Whole exon coverage

• Tumor vs. normal?

• Copy number?

• RNA expression?

• Rearrangements?

• Methylation?

• Cancer genetics is rapidly expanding with high complexity

• Molecular profiling will drive cancer management

• Continued need for higher-throughput cancer genotyping

SummarySummary

• Next gen sequencing poses a huge informatics challenge

• Clinical next generation sequencing is comingis already here

^

MGH Molecular Diagnostics

Dora Dias-Santagata

Kathy Vernovsky

Arjola Cosper

Breton Roussel

Kristin Bergethon

Hannah Stubbs

Vanessa Scialabba

Sara Akhavanfard

Quynh Lam

MGH Cancer Center

Daniel Haber

David Louis

Leif Ellisen

Darrell Borger

ABI/LifeTech

Kevin McKernan

Rosemary Obrien

Steve McLaughlin

Clarence Lee

Nancy Gangemi

Timothy Harkins

Jeremy Stuart

Eric Tsung

Ion Torrent/LifeTechKenny Fan

Jae Han

Ion Torrent/LifeTech

Andrew Felton

Jason Meyers

Maneesh Jain

Simon Cawley

Stuart Davidson

Mike Lelivelt

Questions?Questions?