Impact of Personalized Medicine for the Practice of Hematopathology

Impact of Personalized Medicine for the Practice of Hematopathology


Adam Bagg MDProfessor

Director, HematologyUniversity of Pennsylvania

Kojo S.J. Elenitoba‐Johnson MDHenry C Bryant Endowed Professor Director of Molecular Diagnostics

University of Michigan

Megan S. Lim MD PhDProfessor

Director of HematopathologyUniversity of Michigan 2011 College of American Pathologists. Materials are used with the permission of the faculty.

1

DisclosuresDisclosures

None

2

Course ObjectivesCourse Objectives

• Be familiar with concepts of personalized medicine in the context of hematopathology

• Understand the appropriate diagnostic and prognostic value of specific molecular tests in the evaluation of hematopoietic neoplasms

• Appreciate how novel genetic alterations in myeloid and lymphoid neoplasms impact patient management.

3

AgendaTopic Speaker Time

I. Introduction Lim 5’

II. Principles of personalized medicine and evolution of molecular hematopathology-Overview of commonly used molecular technologies in hematopathology-Update in new technologies including array CGH, next generation sequencing

Elenitoba-Johnson

40’

III. Approach to myeloid neoplasms in the era ofpersonalized medicine-Chronic myeloid leukemia-Acute myeloid leukemia

Bagg 65’

IV. Approach to NHL in the era of personalized medicine-Aggressive B-cell lymphomas-Splenic Lymphoma-ALCL

Lim 60’

IV. Summary and Closing Lim 10’4



Kojo S.J. Elenitoba‐Johnson MDHenry C Bryant Endowed Professor

Director of Molecular Diagnostics LaboratoryUniversity of Michigan

2011 College of American Pathologists. Materials are used with the permission of the faculty.5

Personalized medicine and evolution of molecular hematopathology

1. Principles of personalized medicine

2. Evolution of molecular diagnostics

3. Update in new technologies including array CGH, next-generation sequencing

6

Concepts of Personalized Medicine

Personalized Medicine (2013) 10(1), 13-15

7

Personalized Medicine

DefinitionForm of medicine that uses information about a person’s genes, proteins, and environment to prevent diagnose and treat disease National Cancer Institute http://www.cancer.gov/

Not a new concept

“The new language of genomics, as applied to medicine, is less a revolution than an evolution“ Steele 2009 Personalized medicine: something old, something new. Pers Med 6:1‐5

8

Genetic testing for personalized medicine in hematopathology

9

Pre-symptomatic risk assessment

Diagnosis Prognosis Treatment

Pharmacogenomics

Classification Disease monitoring

Response to treatment

Prediction of response

1980 1985 19901988 1994 2001 2008 …

Morphologic evaluation

Cytogenetics

Introduction of immunopathology, flow cytometry immunohistochemistry

Rapid growth in molecular genetic application

Gleevecapproved by FDA

New therapies

•Combination of small molecules

Clinical sequencing for detection of structural alterations/mutations

2015

Evolution of Molecular Diagnostics in Hematopathology

PCR FFPE tissues REAL classification

WHO classification

10

MOLECULAR HEMATOPATHOLOGY

11

Genetic alterations in hematologic neoplasms

12

Gene rearrangements

Gene Mutations

Gene Mutations

AdditionsAdditions

LossesLosses

Different types of gene rearrangements

13

VDJ/TCR

Homogeneous Heterogeneous

Monoclonal Polyclonal

Physiologic

Diagnosis

Qualitative Quantitative

Chromosomaltranslocations

Yes or No How much

Pathologic

MinimalResidual disease

Diagnosis

t(11;14)t(14;18)t(2;5)

V D J C Germline(lane A)

18 kb

V DJ C DJ rearrangement(lane B)

12 kb

V DJ C VDJ rearrangement(lane C)

21 kb

A B C21 kb

18 kb12 kb

Southern blot

JB1B2 Probe

JB1B2 Probe

JB1B2 Probe

D

D

5´

5´

5´ 3´

3´

3´

B cells undergo immunoglobulin gene rearrangements

14

Detection of clonal rearrangements

15

Southern blot hybridization PCR/gel

electrophoresis PCR/capillarygel electrophoresis

Applications of Molecular Approaches to Clonality Testing

• Benign versus malignant

• Lymphoma versus carcinoma

• Minimal residual disease

• Lineage: B or T

• Subclassification

16

Common Recurrent Chromosomal Aberrations in Human Malignant Lymphoma

Cytogenetic Abnormality Genes Involved Disease Clinical Features Frequency

(%)t(14;18)(q32;q21) IGH@/BCL2 FL Indolent ~90

t(2;18)(p12;q21) IGK@/BCL2 FL Indolent <5

t(3;14)(q27;q32) IGH@/BCL6 FL Indolent ~10

t(11;14)(q13;q32) IGH@/CCND1 MCL Aggressive >90

Trisomy 3 Unknown MZBCL Indolent Variable

t(11;18)(p21;q21) BIRC3(API2)/MALT1 MZBCL Indolent 40

t(1;14)(p22;q32) IGH@/BCL10 MZBCL Indolent *

t(14;18)(q32;q21) IGH@/MALT1 MZBCL Indolent *

t(3;14)(p22;q32) IGH@/FOXP1 MZBCL Indolent *

t(8;14)(q24;q32) IGH@/MYC Burkitt lymphoma Highly Aggressive 75

t(2;8)(p12;q24) IGK@/MYC Burkitt lymphoma Highly Aggressive 15

t(8;22)(q24;q11) IGL@/MYC Burkitt lymphoma Highly Aggressive 10

t(3;14)(q27;q32) IGH@/BCL6 DLBCL Aggressive ~30

t(14;18)(q32;q21) IGH@/BCL2 DLBCL Aggressive 30

Amplification 2p REL PMLBCL Aggressive *

del 13q14 DLEU2/mir-15a/16-1 CLL Indolent 25-50

Trisomy 12 Unknown CLL Indolent 30

t(2;5)(p23;q35) NPM1/ALK ALCL Aggressive ~40

inv 14(q11;q32) TRA@/TCL1 T-PLL Aggressive 75

17

18

Fluorescence in situ hybridization detection of chromosomal translocations

TranslocationNormal

Gene 1probe

Gene 2probe

Gene 1Bprobe

Gene 1Aprobe

FusionAssay

SplittingAssay

Mantle Cell Lymphoma

BCL-1/IGH PCR

Cyclin D1

CD20+CD43+CD23-

19

FISH detecting t(11;14)

Molecular genetics of hematologic neoplasms

Technologies

• Clonal populations• Rearrangements• Chimeric fusions• Mutations• Splice variants

20

• PCR • Cytogenetics• FISH • CGH arrays• SNP arrays• NGS

Genetic events

21

Current role of molecular diagnostics in hematopathology

WHO 2008 Classification

Diagnosis Prognosis Therapy

Traditional “Genomic” Technology: Metaphase Cytogenetics

• Drawbacks– Requires actively dividing cells for analysis– Cannot be performed on archived tissue– Lower limit of resolution of approximately 2,500 kb for

abnormalities

• Benefits– Routinely performed in many centers– Excellent for detecting large numerical and structural

abnormalities– Can detect balanced abnormalities– Allows distinction between different clones

22

Impact of New Genomic Technologies on Hematopathology

High density aCGH/SNP arrays

Next generation sequencing

23

Array Comparative Genomic Hybridization

24

High Density SNP Array

25

aCGH and SNP Array• Benefits

– Allows resolution of ~10 ‐ 50kb, depending upon coverage– Far better than metaphase karyotyping to identify small copy

number changes– Can be performed on archival tissue, with some loss in quality– SNP arrays allow detection of copy‐neutral loss of

heterozygosity (acquired uniparental disomy)

• Drawbacks– Are unable to identify balanced translocations– Identify many benign CNVs—paired normal tissue helpful to

identify somatic changes in tumors

26

BM

CD3

SNP array detection of aquireduniparental disomy

27

PRIVATE SECTOR PUBLIC SECTOR

29

Next generation sequencing

Human Genome Project

Rapid development in cost and quality trends of DNA sequencing

30

Sanger sequencing

PCR followed by cycle sequencing with dNTPs/ddNTPS

Electrophoretic separation of chain termination products

31

Traditional DNA “Sequencing” MethodsMethod Basic Technique Sensitivity Advantages DisadvantagesSanger “chain‐terminationmethod”

Fluorescent dye–labeled bases; DNA fragments separated by capillary electrophoresis

High “Gold standard”; complete sequence

Time‐consuming; cannot detect deletions, translocations, copy number changes

Pyrosequencing“sequencing by synthesis method”

Chemiluminescentdetection; DNA polymerase synthesizes DNA complementary to a target template; pyrophosphate release detected at each base addition

Higher More sensitive than Sanger; provides percentage of mutated vs wild‐type DNA; works well with fragmented DNA from FFPE samples

Short length reads limit technique to hot spots; limited accuracy detecting changes in homopolymer runs; scalability limited com‐pared with other NGS methods

Mass spectroscopy–based mutation analysis (MALDI‐TOF)

Primer extension with chain termination using PCR amplicon templates identifies variant alleles

Higher High sensitivity; high resolution of DNA fragments; detects frame‐shift mutations; readily identifies germline SNPs

Mass spectrometry resolution window balanced with PCR amplicon design requirements combine to limit scalability

Allele‐specific RT‐PCR

Primers span DNA sites of interest; probes detect specific mutations

Higher Very high sensitivity; widely used for clinical testing for oncogene mutations in CRC, NSCLC

Scalability constraints limit application to hot spots

RT‐PCR melting curve analysis

Heterogeneous DNA PCR products melt at temperatures different from those for homogeneous DNA PCR products

High High sensitivity; provides percentage of mutated vs wild‐type DNA

Often difficult to resolve differences in melt curves; difficult to standardize; multiplex capability limited

32

Next‐generation sequencing

• Massively parallel sequencing

• Powerful approach to DNA sequencing

• Dramatic reduction of cost‐per‐base and time

• Disruptive technology resulting in dramatic change in rate of new discoveries

33

NEXT GENERATION SEQUENCING

DNA sequencingTranscriptomesequencing

Gene regulation and control analyses

Whole genome/Exomesequencing/resequencing

De novo sequencing/Targeted

resequencing

SNP, CNV, deletion, Translocations‐Chromosomal

rearrangement

Tag profiling

Small RNA discovery

mRNA seq

Mapping reads Non mapping reads

Gene expression

SNP discovery

Fusion transcripts

Alternate splicing

Viral genome

ChIP Seq

DNA protein interaction

Genome wide methylation

34

Comparison of Sanger and Next‐Generation Sequencing Methods

High‐throughput Sanger• Multiple copies of single

fragment produced• ~100 fragments sequenced

in parallel• Each fragment ~1kb• 0.1 Mb of sequence per

“run”

Next‐generation• Multiple copies of multiple

fragments produced• Millions of fragments

sequenced in parallel• Each fragment 50‐100bp• >100 Gb of sequence per

“run”

35N Engl J Med. 2010 May 27;362(21):2001-11

A number of technical platformsbut similar strategy

Miniaturization of individual sequencing chemical reactions to:

– overcome limited scalability of Sanger sequencing

– overcome bottlenecks of library preparation and template preparation

– allows millions of individual sequencing reactions to occur in parallel

– results are high volumes of short read sequence data in unprecedented detail and single‐nucleotide resolution

36

Next Generation Sequencingsequence DNA fragment library in situ

A T C G

Massively paralleled configuration of sequences

37

Read= A short length of sequence

Example: Illumina Hi‐ Seq

• Each read is 100 bp

• 160M of these reads!

• Need to connect sequences to each other

38

Disruptive Technology in Drivers of Medical Genomics

Time Period Genomes Turn‐around time FTEs Cost per genome

1990‐2003 1.NIH reference ~5 years ~2,000 ~$2‐3 billion2.Celerareference

2003‐2009 ~10 additional ~6 months Dozens $300,000→38,000

2010‐2014 103‐105 4 weeks 3‐4 $ 6,000 exome$ 9,500 genome

2015‐2020 Millions 6‐0.5 hr <<1 $1000‐250

39

Whole Genome Sequencing

Human genome = 3.1 billion base‐pairs

WGS – determining the sequence of an individuals genome

Includes sequence of the genes – exons & introns

Includes sequence of regions between genes

40

Adapter ligation

Flow cellcDNA Fragmented

cDNA

QC analysis

mRNA

Whole Transcriptome Sequencing

Paired end sequencing

41

RNAseq

Exome sequencing: sequencing the coding region of all genes

• Human Genome: 20K to 25K genes

• Genes composed of exons that code for AA of a protein

• Introns are spacer regions that are spliced out

• Can interpret a change in AA sequence such as an Arg to a stop codon

42

Whole Exome Sequencing

• Exons are shown as colored part of gene

• Capture array has complementary sequence of each exon bound to solid support

• Single strand DNA of exons hybridize • Selected DNA sequenced • 1% of the genome

43

WES vs WGS vs RNAseq

Whole exome Whole genome Whole transcriptome

Coverage of genome 1% Better coverage mRNA, tRNA,

Cost* $600‐800 $1000‐3000 $700‐800

Genome Only coding region

Coding and noncoding mRNA, tRNA, alternativetranscripts

Bioinformatics Better developed Challenging Better developed

Applications SNV, indels Rearrangements, SNV, indels

Rearrangements, inversions, SNV, indels,GEP, alternative spliced variants

44

Commercial Next‐Generation Sequencing Platforms

Technologic Basis Application Time RequiredInstrument Cost($, thousands)

Illumina Flow cell‐based, reversible dye termination, and4‐color optical imaging

Whole exomesequencing; whole genome sequencing; SNP detection

Intermediate, 4 d fragment; 9 dpaired end

500‐900

454 Pyrosequencing

Emulsion PCR with bead‐based pyrosequencing and CCD light imaging

Targeted exon sequencing; confirmatory sequencing; SNP detection

Fast, <1 d 500‐700

Helicos Oligo‐dT captured PolyA‐tailed DNA fragments; flow cell 4‐color dNTP optical imaging

Single molecule sequencing; whole genome sequencing

Slowest, 8 d 999

SOLiD Sequential dinucleotide ligation; flow cell‐based 4‐color optical imaging

Whole exome sequencing; whole genome sequencing; SNP detection

Slow, 7 d fragment run; 14 d paired end

600‐700

Ion Torrent Semiconductor‐based nonoptical detection; standard dNTP sequencing chemistry

Targeted sequencing projects not demanding deep sequencing

Fast, <1 d 50

45

Performance characteristics of NGS platforms

100 1000 10000 100000

1 Mb

100 Mb

1 Gb

10 Gb

100 Gb

Read Length bp

Total datayield

Sanger

Roche454 PacBioRS

Ion Torrent

IlluminaHiSeq2000

Ion Proton

IlluminaHiSeq2500

White = frozen, orange = FFPE

MR Schweiger et al. PLoS ONE 2009;4:e5548 M Kerick et al. BMC Medical Genomics 2011;4:68

Targeted high throughput sequencing of FFPE tissuesis feasible

47

Novel alterations in lymphoid neoplasia identified by NGS

• Chronic lymphocytic leukemia

• Splenic marginal zone lymphoma

• Hairy cell leukemia

• Follicular lymphoma

• Diffuse large B‐cell lymphoma

• Burkitt lymphoma

• Peripheral T cell lymphoma, NOS

• Angioimmunoblastic T cell lymphoma

• Cutaneous anaplastic large cell lymphoma

• Clonal evolution

• Tumor heterogeneity

48

2015

Bioinformatics Analysis of Genome Sequencing Data

The steps from sample processing to result interpretation

Sample

DNA

FASTQ

BAM

VCF

aVCF

Interpretation

ATGACGATGATAGGATGA

chr19: 17,949,097 ATGACGATATAAGATGA

chr19: 17,949,108 G A Ref: 35 Alt:39 GQ: 99

chr19: 17,949,108 G A JAK3 p.M511I MISSENSE NM_000215

Known pathogenic mutation in T-cell leukemia with targeted therapeutic

What are the barriers to automated genome sequencing interpretation?

Genotype-phenotype correlations are known for 500+ conditions

Increasingly, these data are available in public databases…

True Positives and

True Negatives

Diagnosis

Genome Sequencing DataThe challenges associated with automated interpretation

Negative Positive

False FN FP

True TN TP

False Negativespotentially missed diagnoses

True Negativeswhat can safely be ignored

False Positivestoo much to look at

True Positiveswhat needs to be acted on

Sensitivity how often positive when it should be?

Specificity how often negative when it should be?

Positive Predictive Value how often positive if test result is positive?

Negative Predictive Value how often negative if test result is negative?

Practical Aspects of NGS in the Clinical Laboratory

Rigorous Validation Required!

• Verify all secondary mutations identified during validation using known standards

• Run normals – Tissues and HAPMAP cell line

• Orthogonal and parallel platform validations

• Bioinformatics Validation

• Variant database construction

• External Proficiency and User Community Groups

Targeted Sequencing Assays

Recommendations for Development and Validation

Ion Torrent 50 gene panel – depth & uniformity

8 specimens on 316v2

8 specimens on 318v2

Ion Variant Caller Report

Mixed Mutation Control

Cell block mix of 4 engineered cell linesWe extract DNA from scrollsCost effectiveMutation levels determined/verified via digital droplet PCR

KRAS – G13D (16.5%) and G12D (5.5%)FFPE reference cell line mixture; mutations quantified via digital PCR8 specimens on Ion Torrent 316v2 chip

KRAS G12D

KRAS G13D

KRAS – G12D low‐level

KRAS G12D

FFPE tissue sequenced via Sanger at ~7% mutation8 specimens on Ion Torrent 318v2 chip

MPL W515L (low‐level)

MPL W515L

Peripheral blood tested positive for W515L via allele‐specific PCRTested negative via Sanger sequencing

DL‐12‐16233

Conclusions

• NGS is a disruptive technology

• NGS is transforming the routine practice in molecular pathology and personalized diagnostics

• Robust bioinformatics infrastructure is essential in NGS implementation

• Pathologists should embrace this platform for integration in all aspects of Diagnostic Pathology

QUESTIONS?



II. Principles of personalized medicine and evolution of molecular hematopathology-Overview of commonly used molecular tests in hematopathology-Update in new technologies including array CGH, next generation sequencing

Elenitoba-Johnson

40’


Bagg 65’


Lim 60’



62

Approach to myeloid neoplasms in the era of personalized medicine

Adam Bagg MDUniversity of Pennsylvania

Case #1

30-year-old previously healthy female patient

- nonspecific abdominal pain - noted to have a leukocytosis on a routine CBC- a bone marrow biopsy was performed

- digital whole slide images of PB and BM Bx

• What is the favored diagnosis based upon the clinical

presentation and peripheral blood smear morphology?

What is the role of performing a bone marrow biopsy in

this case?

• What additional studies would you perform to confirm

the diagnosis?

Case #1 – Questions

• The patient read on the internet that the presence of

ABL1 mutations might affect her response to therapy,

and wants to be tested for these; should she be

tested now? What are the indications for testing for thesemutations?

• How should the patient be monitored after initiation of

therapy?


CML - a peripheral blood diagnosis

Major flavors of myeloid neoplasms

AML

MPN

MDS

MPN/MDS

MALNWEAAOPPOF

Myeloproliferative -vs- Myelodysplastic

bone marrowhypercellularity

MPNhigh

peripheralcounts

MDSlow

peripheralcounts

quantitativelyincreased

hematopoiesis

effective

ineffective

neoplastichematopoietic

stem celldisorders

CEL Mastocytosis

UnclassifiableCNL

Flavors of myeloid neoplasms (non-AML, non-MDS)

CML

Ph-neg MPNs“Classical”

“Non-classical”

MPN

CMML

JMML

aCML

Unclassifiable

PDGFRA PDGFRB FGFR1

MPN/MDS

MALNWEAAOPPOF

PV

ET

PMF

Mast cell

SM

MALNWEAAOPPOF

CML

PV ET PMF

BCR-ABL1

JAK2JAK2

MPL

KIT

PDGFRAPDGFRBFGFR1

CEL

CNL

CMML JMML aCML

TET2ASXL1

SRSF2

PTPN11

RASNF1

CALR

CSF3R

SETBP1

The (other) Philadelphia story

• median age ~ 67 years

• non-specific symptoms- fatigue/weight loss/LUQ discomfort- routine CBC (~40%)

• splenomegaly (~90%)

• is a peripheral blood diagnosis [but defined by genetics]- marked leukocytosis (> 50 -100 x 109/l) - granulocytes in all stages of maturation (but < 10% blasts)- [need to rule out a “leukemoid reaction”]- basophils- platelets , N or - hemoglobin mildly

Chronic myelogenous leukemia

Blast phase

CML is (was) a triphasic disease

Accelerated phase

Chronic phase

Genetic testing in CML

Diagnosis

Monitoring

Resistance

Diagnosis

Monitoring

Resistance

? CML cytogenetics

CML ~2.5% -~2.5% + CML

Avoid term“Ph-neg” CML

(no such thing!)

? CMML? aCML

? BCR-ABL1

~95% +

CML

~5% -

moleculargenetics

? t(9;22)

CML diagnosis: t(9;22) and BCR-ABL1

? CML cytogenetics

CML ~2.5% +

? moleculargenetics

YES!molecular target for:

1] Rx2] MRD

? cytogenetics

YES!- clonal “evolution”

[Ph+; Ph- with imatinib]

? BCR-ABL1

~95% +

CML

~5% -

moleculargenetics

? t(9;22)

CML diagnosis: t(9;22) and BCR-ABL1

Woessner et aCancer J 201117; 477-486

ROS

Genomic Instability

Progression

TKDmutations

Monitoring

Molecular testing in CML

Diagnosis

Monitoring

ResistanceResistance

CML monitoring

Two major forms of therapy …

TKI

SCT

• Initial therapy of choice• Does not eradicate/cure CML (?)• ? Long-term outcome• Minimal toxicity1

1 cytopenias, GI/pancreas, fluid retention, myalgia

• No longer 1st line Rx2

• Only Rx that cures CML3

• Major toxicity and mortality4

• Not available to all5

BCR-ABL1reduction

BCR-ABL1negativity

Rx goal (?)

2 indicated when [i] very young, [ii] (selected)TKI failure, [iii] AP an3 10-yr survival ~65%4 10-20% mortality even when low risk 5 need to have a donor!

CML monitoring

sensitivities…modalities…

CBC ~10% [~10-1]

conventional cytogenetics

~5% [~10-2]

D-FISH ~0.5% [~10-3]

RT-PCR [quantitative] ~0.001% [~10-5]

RQ-PCR [qualitative] ~0.001%* [~10-5]* non-nested; can reach 10-8 with nested RT-PC

CML monitoring: definitions of response

complete hematologic

• platelet: < 450• WBC: < 10• diff: no immature granulocytes• basos: < 5%• clinical: non-palpable spleen

cytogenetic # Ph+

• none: >95% • minimal: 66-95%• minor: 36-65% • partial: 1-35% • complete: 0%

molecular

• next slide please …{major

diagnosis

complete hematologic remission

complete cytogenetic remission

major molecular response(MR3.0)

“complete molecular remission”undetectable transcript

(MR4.5)

1012

<1010

<109

<107-8

100

<0.1

<0.01

responses BCR-ABL1ratio

… yet overall survival ~85%

imatinibresponses

~98%

~85%

~75%

~10%

logreduction

<1011

>3

# cells

24% @ 6 mo39% @ 1 yr55% @ 2 yr86% @ 8 yr

? up to 40-50%800mg vs 400mg

but imatinib failure-free survival ~60% (SE, LOR)

The sooner the better …

• predicting adverse outcomes if don’t reach landmarks:- 18 months <3 log reduction (> 0.1%) = MMR- 12 months <2 log reduction (>1%)- 6 months <1 log reduction (>10%)

• likelihood of an adverse event, if achieve MMR by:- 12-18 months ~15%

- 6-12 months ~8%

- 6 months ~0%

• thus, the quicker the response, the better the outcome

- testing as early as 3 months (or even 1 month) may be predictive!level >10% at 3 months (ie, < 1 log reduction) particularly poor

RQ-PCR: Ongoing attempts at harmonization

international scale

conversionfactor

internationalstandard

• standardized baseline from original IRIS trial• pool of 30 CML cases

• sample exchange between your lab and IRIS lab• analogous to INR

• WHO initiative—freeze dried preparations • 4 serial dilutions of K562 into HL60:

10%, 1%, 0.1%, 0.01% • also insufficient amounts of this!• use restricted to:

reference laboratoriesmanufacturers of secondary reference materi

rapidly exhausted

Molecular testing in CML

Diagnosis

Monitoring

Resistance

Resistance to imatinib (and other TKIs …)

primary • failure to achieve an initial response

• uncommon--seen in ~5% of pts

• bioavailability; not due to mutations

secondary • commoner--usually due to mutations

• response loss of response (especially early on)- 1st 3 years on Rx: ~5%/year

- 2nd 3 years on Rx: ~1%/year

• hence 1st 3 years most important monitoring perio- most likely to encounter mutations

• incidence of mutations: depends on disease phas

Mechanisms of resistance to imatinib

BCR-ABL1independent

BCR-ABL1dependent

1. Kinase domain mutations:- most common cause of resistance [~60% overall]

- spans ~240 aa’s

2. BCR-ABL1 amplification:- genomic > transcriptional [~10%]

3. Clonal evolution:- other genetic/cellular pathways [LYN/JAK-

STAT/SRC/GAB2/PI3K]4. ↓Bioavailability:- absorption- metabolism [hepatic]- plasma binding [1acid glycoprotein sequestration]- ↓influx ↑efflux [MDR1, PGP, BCRP2/ABCG2, hOCT1,

MRP1]5. 7. BIM6. TKI BCL6

T315I

Kinase domain mutations

P = P loop- ATP-binding site- ? worst mutations

B = Binding domain- where imatinib binds

C = Catalytic doma

A = Activation loop- conformation altered- affects imatinib bindi- closed: inactive- open: active

2-10% of patients

>10% of patientsgreen

red

> 100 different mutations[ABL1 exons 4-10]

these 6 accountfor > ~65% ofall mutations

~15-20%

T315I

Kinase domain mutations

• MUST test for specific mutations before changing Rx to

determine appropriate action

1. do nothing … may be innocuous (SNPs)2. increase dose of imatinib3. switch to 2nd (or 3rd) generation kinases

inhibitor ** type of mutation will dictate which one to use

4. stem cell transplant5. alternative Rx/clinical trail

• If a mutation is found, options include:

• If a mutation is NOT found:1. consider compliance failure 2. if due to other biologic causes problematic

Indication for mutation testing

treatment failure/suboptimal response

loss of response

anyone in AP or BP

• no CHR by 3 months• no PaCR by 6 months• no CCR by 12 months• no MMR by 18 months

• hematologic relapse• cytogenetic relapse• ↑ing BCR-ABL1 levels [1-log]

No role at CP diagnosis* NEVER change Rx based upon a single result:(1) short-term lapse in compliance(2) assay variation

P h l i i ti ith l f

*

• BCR/ABL1 >10% at 3 month

Methods of mutation testing

Technology Sensitivity

Specificity Bias

Sanger sequencing 15-25 ++ no

Subcloning and sequencing 10 +++ noD-HPLC 0.1-10 ++ noPyrosequencing 5 ++ noDouble-gradient denat. electroph.

5 ++ no

Fluorescence PCR – PNA clamping

0.2 ++ yes

ASO-PCR 0.01 ++ yesSmall clones (<20%) may not be clinically signific

Sanger sequencing (current) method of chto detect, then pyrosequencing to quantify

Emerging data to support more sensitive detectio if have T315I @ 10-5 precludes MMR!

High resolution meltingNanofluidicsNext gen. sequencingMass Spec

Overcoming resistance

• Alternative kinase inhibitors:- 2nd generation TKIs: nilotinib- dual SRC-ABL inhibitors: dasatinib

bosutinib - multikinase (KIT/PDGFRA/FLT3/FGFR1) inhibs: ponatinib- aurora kinase inhibitors: XL-228

MK-0457

• Others:- protein synthesis inhibitors: omacetaxine- switch pocket-control inhibitors: rebastinib (DCC-2036)

• ImmunoRx:- vaccines against: BCR-ABL1, PR1, WT1, HSPs- 1o role in MRD Rx

• Targeting the stem cell:- not dependent on BCR-ABL1 WNT-CTN/HH/JAK2-PP2A/TGF-

**

* FDA approved

**

∞ active vs T315I

∞

∞

∞

∞

∞

Overcoming resistance

• Alternative kinase inhibitors:- 2nd generation TKIs: nilotinib- dual SRC-ABL inhibitors: dasatinib

bosutinib - multikinase (KIT/PDGFRA/FLT3/FGFR1) inhibs: ponatinib- aurora kinase inhibitors: XL-228

MK-0457

• Others:- protein synthesis inhibitors: omacetaxine- switch pocket-control inhibitors: rebastinib (DCC-2036)

• ImmunoRx:- vaccines against: BCR-ABL1, PR1, WT1, HSPs- 1o role in MRD Rx

• Targeting the stem cell:- not dependent on BCR-ABL1 WNT-CTN/HH/JAK2-PP2A/TGF-

**

* FDA approved

**

∞ active vs T3151

∞

∞

∞

∞

∞® ~20% vascular/

thrombotic problems®

80 -

100 -

60 -

20 -

40 -

0 -

CCR (1 yr) MMR (1 yr)

66%

28%

77%

46%

CCR (1 yr) MMR (1 yr)

65%

22%

80%

44%

2nd generation TKIs as 1st line therapy

imatinib 400mg

dasatinib 100mg

imatinib 400mg

nilotinib 300mg x 2

Genetic testing in CML: summary

Diagnosis

Monitoring

Resistance

CC BM PB FISH instead is not ideal

RT-PCR PB qualitative (not quantitative) with characterization

CC BM 3-6 monthly until CCR6-12 monthly thereafter

FISH PB ? before achieve CCR (not ideal)

RQ-PCR PB 3 monthly until MMR, then 6 monthly (ELN)

directsequencing

PB vBM

Rx failure, loss of response, accelerated & blast phase

Survival in newly-diagnosed CP CML by year of Rx

<19751975-1982

>2001

1991-2000

1983-1990

Kantarjian H et al. Blood 2012;119:1981-1987

• What is the favored diagnosis based upon the clinical

presentation and peripheral blood smear morphology?

What is the role of performing a bone marrow biopsy in

this case?

• What additional studies would you perform to confirm

Case #1 – Answers

-- chronic myelogenous leukemia, BCR-ABL1positive

-- best shot at getting metaphases

-- qualitative RT-PCR (FISH if negative or complex)

• The patient read on the internet that the presence of

ABL1 mutations might affect her response to therapy,

and wants to be tested for these; should she be

tested now?

• Would you be surprised if cytogenetics by metaphase

analysis showed a normal karyotype?-- a little (~ 2-3% missed)

-- probably not, but things may change

Case #1 – Answers

What are the indications for testing for thesemutations?

• How should the patient be monitored after initiation of

therapy?

Case #1 – Answers

-- failure to reach milestones-- loss of response-- accelerated phase or blast phase

-- CBC-- BM cytogenetics-- RQ-PCR frequency depends on attainment of

milestones

• How will the finding of a mutation affect subsequent

therapy?

Case #1 – Answers

-- do nothing … may be innocuous-- increase dose of imatinib-- switch to 2nd (or 3rd) generation kinases inhibitor

** type of mutation will dictate which one to use-- stem cell transplant-- alternative Rx/clinical trail

Case #2

57-year-old male patient--recent onset of bleeding gums

- CBC: HGB 10.2 g/dLWBC 7.4 x 109/L PLT 21 x 109/L

- abnormal cells (79 %) on peripheral blood sme- bone marrow aspirate/biopsy performed

- digital whole slide images of PB and BM Bx

• What is the favored diagnosis based upon the

clinical presentation and morphology?

• What standard additional studies would youperform to confirm and refine the

diagnosis, andwhich of these are likely to have the

greatestimpact on prognosis and therapy?


• Would you be surprised if cytogenetics bymetaphase analysis showed a normalkaryotype?


• How might the finding of a normal karyotype

affect what additional genetic tests you might

order?• In addition to the molecular studies ordered as

per questions 2 and 4 above, what other genetic

analyses are likely to be useful in terms ofprognosis and therapy?• What technology might be best to test for

allthese mutations?

Acute myeloid leukemia

morphology

cytochemistry

immunophenotype

the most important diagnostic test inacute leukemia

(cyto)genetics

Acute leukemia

Risk group Genetics Frequency CR 5yr Sur

t(8;21)Favorable t(15;17) ~25% ~85% ~60%

inv(16)

normalIntermediate +8, +21 ~50% ~80% ~40%

t(9;11)

-5, -7Unfavorable inv(3)

complex ~25% ~60% ~15%monosomal11q23t(6;9)

M2M3M4Eo

M5

dysmeg

basophilia

AML: Cytogenetic stratification


100

60

80

40

20

1 32 4 5 6 7 years

surv

ival

intermediate

unfavorable

favorable

AML: Cytogenetics

The big 4 sine qua non for WHO

• t(15;17) [M3] ~10%• t(8;21) [M2] ~10% GOOD• inv(16) [M4Eo] ~10%

• t(11q23)* [M4/M5] ~5 - 10% not so go

© 2001 2008 = big 7 [added t(1;22), t(6;9) and inv(3); t(9;11) specifically]

* promiscuous (>60 partners)

}}

AML: Predicting the genetics

Translocation Partner Frequency ATRA-responsive

t(15;17)(q24;q21) PML >99% +t(11;17)(q23;q21) ZBTB16 0.5% -t(11;17)(q13;q21) NUMA <0.5% ?+t(5;17)(q32;q21) NPM <0.5% ?+t(17;17)(q21;q21) STAT5B <0.5% ?

The various genetic flavors of APL

Non-t(15;17)-positive cases not “APL”

Another morphologic flavor of APL

AML with t(8;21)(q22;q22); RUNX1-RUNX1T1

AML with inv(16)(p13.1q22) or t(16;16)(p13.1;q22); CBFB-MYH11

AML with t(6;9)(p23;q34); DEK-NUP214

AML with inv(3)(q21q26.2) or t(3;3)(q21;q26.2); RPN1-EVI1

AML: Cryptic abnormalities

t(8;21)RUNX1-RUNX1T1

inv(16)CBFB-MYH11

t(15;17)PML-RARA

• ~ 30% of positive cases lack M4Eo morphology• upto 30% with M4Eo morphology are:

- RT-PCR/FISH- cytogenetics

• some M4Eos do not have inv(16) at all

• upto 30% are:- RT-PCR/FISH- cytogenetics

• upto 15% are:- RT-PCR/FISH

t ti

+-

+-

+-

Mutations in AML

15 1614 20191817

ECD JM TK2 TM

*

400bp

300bp

350bp

FLT3ITD heterozygous FLT3ITD hemizygous

+ -FLT3WT

FLT3 : genotyping

FLT3 : multiplex PCR on CE

ITDe14/e15

e20/ECoRV D835

• One of the most common known molecular targets in

• Possibly most prognostically relevant molecular lesion- quantification is important (mutant:wt)

• A potentially useful MRD target- patient specific primers vs ITD- ITD may [rarely] not be stable between presentation and re

[+ -] [- +] [+A +B] [+ ++]

• A therapeutic target- a la imatinib mesylate in CML

FLT3

NPM1

• most commonly mutated gene in AML- ~60% of cytogenetically normal AML- ~30% of all AML- shuttles between nucleus and cytoplasm (lives in nucleus)- mutations (typically 4bp insertion):

* lose nucleolar localization signal/creates export signal

- good prognosis:* but only if FLT3 wild type (and perhaps if IDH2 mutant?)

without IHC IHC—wild type IHC—mutant

AML: Genetics – WHO 2008

Karyotype

Genes Morphology

Frequency

Prognosis

t(8;21) RUNX1/RUNX1T1

M2 ~5% good

t(15;17) PML/RARA M3 ~5-8% good

inv(16) CBFB/MYH11 M4Eo ~5-8% good

t(9;11) MLLT3/KMT2A M4/M5 ~2% intermediate

t(1;22) RBM15/MKL1 M7 <1% good

t(6;9) DEK/NUP214 Basophila ~1% poor

inv(3) RPN1/EVI1 MLD ~1% poorNPM1mutations*

M4/M5 ~30% good

CEBPAmutations*

M1/M2 ~10% good* AMLs with mutated NPM1 and CEBPA = provisional entitiesAMLs with FLT3 mutations not an “entity” but testing “strongly recommend

AML genetics – but wait, there’s more…

• explosion of discoveries (NGS, SNPa …)

• enriched in cytogenetically normal (CN) AML• unlike most “AML-defining” translocations:

- may be (variably) seen in other myeloid malignancies

(MDS, MPN, MDS/MPN, 20 AML)• relevant in prognosis may dictate Rx

AML mutations: the NKOTB

TET2

CBL

DNMT3A

IDH1/2

EZH2ASXL1

MLL

WT1

RUNX1SF3B1BCOR

PTPN11

JAK3

ND4

NPM1

CEBPA

FLT3

KIT

Meme K K

me K

hmeC

me

me

IDH1/2

IDH1/2

MLL

EZH2

DNMT3

TET2

gain-of-function loss-of-function

me

ASXL1KMT2A

AML

AML: Cooperative mutations

FLT3RASKIT

PTPN11JAK2

Class I mutations

Signal transducers:• [+] proliferation• [+] survival

Class II mutations

PML-RARARUNX1-RUNX1T1

CBFB-MYH11KMT2A fusions

?NPM1

Transcription factors:• [-] differentiation• [-] apoptosis

Class III mutations

Epigenetics

IDH1/2TET

DNMT3A

AML: But wait, there’s more … 99.5% of cases

Activated signalling

DNA methylation

NPM1 mutation

Chromatin modifiers

Transcription factor fusions

Transcription factor mutations

Tumor suppressors

Cohesin complex

Spliceosome

FLT3 RASs PTPs

IDH1/2 TET2 DNMT2A

KMT2A ASXL1 EZH2

PML-RARA CBFB-MYH11

RUNX1 CEBPA

TP53 WT1 PHF6

RAD21 SMC3 STAG2

SF3B1 SRSF2 U2AF1

~60%

~45%

~30%

~30%

~20%

~20%

~15%

~15%

~15%

AML with inv(16)(p13.1q22) or t(16;16)(p13.1;q22); CBFB-MYH11

+8mKIT

FLT3-TKD

+22

mRAS

15%

35%

15%

20%

50%

}

AML with FLT3 or NPM1 or IDH1/2 mutation

AML: how to manage the nascent data

• has become extraordinarily complex

• reaching the point of TMI

• discovery outpacing clinical trials

• not all are mutually exclusive- some associations and interactions

• how to handle?- algorithms- multiplexing

AML: associations and interactions

Patel JP et al. Integrated genetic profiling in acute myeloid leukemia. N Engl J Med, 2012;366:1079 1089


Risk group Genetics Frequency CR 5yr Sur

t(8;21)Favorable t(15;17) ~25% ~85% ~60%

inv(16)

normalIntermediate +8, +21 ~50% ~80% ~40%

t(9;11)

-5, -7Unfavorable inv(3)

complex ~25% ~60% ~15%monosomal11q23t(6;9)

Normalcytogenetics

Cytogeneticallyintermediateprognostic

group

mutation

NPM1 But only if FLT3 WT(and ? IDH2 mutated)

CEBPA But only if biallelic

FLT3 But only if ITD and no WTAnd no effect on t(15;17) TET2

But only if CEBPA WTAnd only if R882But benefit from hi dose dauna

DNMT3A

PHF6

ASXL1

AML: Cytogenetic molecular stratification

• What is the favored diagnosis based upon the

clinical presentation and morphology?-- acute leukemia

• What standard additional studies would youperform to confirm and refine the

diagnosis, andwhich of these are likely to have the

greatestimpact on prognosis and therapy?

-- flow cytometry-- cytochemistry [?]

Case #2 – Answers

Case #2 – Answers

• Would you be surprised if cytogenetics bymetaphase analysis showed a normalkaryotype?

-- no, they are normal in ~45%

• How might the finding of a normal karyotype

affect what additional genetic tests you might

order?-- FISH or RT-PCR for recurrent,

disease

Case #2 – Answers

• In addition to the molecular studies ordered as per

questions 2 and 4 above, what other genetic analyses are likely to be useful in terms ofprognosis and therapy?

-- mutational analysis of ~10-30 other genes [IDH1, IDH2, TET2, ASXL1, DNMT3A, RUNX1,

KIT, TP53, NRAS]

• What technology might be best to test for all these

mutations?

QUESTIONS?



II. Principles of personalized medicine and evolution of molecular hematopathology-Overview of commonly used molecular tests in hematopathology-Update in new technologies including array CGH, next generation sequencing

Elenitoba-Johnson

40’


Bagg 65’


Lim 60’



Approach to non-Hodgkin lymphoma in the era of personalized medicine

Aggressive B-cell lymphomasSplenic lymphoma

Anaplastic large cell lymphoma

Megan S. Lim MD PhDUniversity of Michigan

143

Case 3

• A 40 year old woman presented with enlarging neck lesion. The biopsy of the right neck lymph node revealed a soft tan lymph node with dimensions of 3.2 x 1.6 x 0.6 cm.

• An initial immunohistochemical evaluation reveals a B cell lymphoma (CD20 and CD10 positive). A virtual slide of the sections of the lymph node is provided.

144

Questions • What are the differential diagnostic considerations?• What additional stains would you perform to refine the

diagnosis?• What additional molecular studies may help refine the

diagnosis?• What is the role of karyotypic analysis, FISH and gene

rearrangement in the workup of this class of lymphoma? • What other lymphoid neoplasms demonstrate MYC

aberrations? • What genetic abnormalities have been recently described

in the group of aggressive B‐cell lymphomas?

145

Case 3: H&E

146

BCL2 BCL6

CD3

CD10

CD20 Ki67

Diagnosis ?

Diffusely + Background T cells >95%

Faintly variably + Diffusely +Small subset +147

WHO classification 2008Aggressive B‐cell Lymphoma

• Diffuse large B‐cell lymphoma, not otherwise specified

• Diffuse large B‐cell lymphoma, subtypesT‐cell/histiocyte‐rich large B‐cell lymphoma• Primary mediastinal (thymic) large B‐cell lymphoma• Intravascular large B‐cell lymphoma• DLBCL associated with chronic inflammation (New entity)• ALK‐positive B‐cell lymphoma• Plasmablastic lymphoma• Large B‐cell lymphoma arising in HHV8‐associated with Castleman disease• Primary effusion lymphoma

• Burkitt lymphoma• B‐cell lymphoma, unclassifiable, with features between DLBCL and Burkitt lymphoma

148

Differential Diagnosis of Aggressive B cell Lymphomas

Histology Immunophenotype Proliferative Index

MolecularGenetics

DLBCL Large centroblasticcells

CD5+/-, CD10+/-CD20+, BCL2+, BCL6+, CCND1-

<90% t(14;18) ~20%, t MYC (10%) 3q27 (~20%)

Burkittlympoma

Intermediate monotonous cells multiple nucleoli

CD5-, CD10+, CD20+, BCL2-,

BCL6+, CCND1 -

>95% t MYC ~100%

Blastoid mantle cell lymphoma

Intermediate lymphoblast like cells

CD5+, CD10-, CD20+, BCL2+, BCL6-, CCND1+

<90% t(11;14)~100%

B-cell lymphoma, unclassifiable, with

features between DLBCL and Burkitt

lymphoma

Intermediate CD5-, CD10+, CD20+, BCL2-,

BCL6+, CCND1 -

>90-95 t(14;18) ~20%, t MYC (10%) 3q27 (~20%)

149

Classic Burkitt lymphoma

150

EndemicSporadicImmunodeficiency

associated

MYC rearrangements in BL

1518q24

8q24

t(8;14)/t(8;22)/t(2;8) low complexity genotype; no coexisting translocations involving bcl2 or bcl6 or bcl1

B‐cell lymphoma, unclassifiable, with features between DLBCL and Burkitt

lymphoma

• 4% of NHLs• Extranodal disease• High proliferative index• Low median survival• Ig/Myc rearrangements or variant myc partners • Includes “double‐hit” lymphomas

152

Double Hit Lymphomas

• Aggressive B‐cell lymphomas harboring MYC/8q24 gene rearrangement with another recurrent breakpoint

• BCL2/IGH + MYC/8q24 DHL 60%• BCL6/IGH + MYC/8q24 DHL 8%• CCND1/IGH + MYC/8q24 DHL 10%• BCL2/IGH + BCL6/IGH DHL Rare

153 Aukema SM et al., BLOOD 2010

DLBCL BL BLU

Age at presentation Usually older but can occur at any age

Children, young adults Older adults

PathogenesisMay be related to the germinal center (GCB), activated B cell or other pathway

GCB derived GCB derived

Growth rate RapidExtremely rapid, Ki67 approaching 100%

Extremely rapid but usually less than 100%

Stage Even distribution, 50% stage 1 or 2

Usually high stage Usually advanced III/IV

Bone marrow involvement

Uncommon, often terminal Common Common

CNS involvement UnusualLeptomeningeal disease common at presentation in children and adults60

Common

EBVUncommon in the absence of immunodeficiency or age‐related senescence

>90% in endemic BL 40% in sporadic and HIV‐related BL

Negative

MYC translocationUncommon, usually a secondary event associated with a complex karyotype

Almost always present as initiating event and single abnormality (MYC simple)

Often double hits with translocations involving MYC, plus BCL2 and/or sometimes BCL6

DLBCL, BL and B-cell lymphoma unclassifiable, with features intermediate between DLBCL and

BL (BLU)

Modern Pathology (2013) 26, S42–S56;154

Is it important to make the distinction between DLBCL, BL and DHL?

155

MYC positive diffuse large B‐cell lymphoma and poor survival

156

CD20

Savage et al. Blood 2009

Ki-67

“Double Hit” Lymphoma

Snuderel et al. AJSP 2010

BL

DLBCLDHL

BACK TO OUR CASE

158

Diagnostic Algorithm

Aggressive B‐cell lymphoma FISH panel• t(14;18) BCL2/IgH

• 3q27 rearrangement BCL6/IgH

• 8q24 rearrangement CMYC/IgH

• t(11;14) BCL1/IgH

159

DDX of BL or BCLU

HistologyImmunohistochemistry

Normal

2G2R

Follicular Lymphomat(14;18) IGH/BCL2

GRFF

BCL2IGH

IGH/BCL2 dual‐color dual‐fusion FISH

160

Abbott MYC Dual Color Break Apart Probe

161

Normal MYC rearrangement

Breakpoint Region

Abbott Molecular

Normal cell (FF) BCL6 translocation +VE (FGR)

telChr 3q27 cenBCL6 gene

TranslocationBreakpoints

Red Probe Green Probe

BCL6 Break-Apart FISH

Integrated interpretation

163

Diagnosis: B-cell lymphoma, unclassifiable, with features intermediate between DLBCL and Burkitt lymphoma (Double-hit)

Aggressive B-cell lymphoma FISH panel• 8q24 rearrangement MYC/IgH +• 3q27 rearrangement BCL6/IgH +• t(14;18) BCL2/IgH -• t(11;14) BCL1/IgH -


164

Diagnosis: B-cell lymphoma, unclassifiable, with features intermediate between DLBCL and Burkitt lymphoma (Double-hit)

Aggressive B-cell lymphoma FISH panel• 8q24 rearrangement MYC/IgH +• 3q27 rearrangement BCL6/IgH -• t(14;18) BCL2/IgH +• t(11;14) BCL1/IgH -


165

Diagnosis: Mantle cell lymphoma, t(11;14) positive with MYC rearrangement

Aggressive B-cell lymphoma FISH panel• 8q24 rearrangement MYC/IgH +• 3q27 rearrangement BCL6/IgH -• t(14;18) BCL2/IgH -• t(11;14) BCL1/IgH +


166

Diagnosis: Burkitt lymphoma

Aggressive B-cell lymphoma FISH panel• 8q24 rearrangement MYC/IgH +• 3q27 rearrangement BCL6/IgH -• t(14;18) BCL2/IgH -• t(11;14) BCL1/IgH -

BCL6‐breakapart FISH

167

C‐MYC (8q24) breakapart probe

168

Integrated interpretation (Diagnosis)

• 8q24 rearrangement (MYC) –Positive

• 3q27 rearrangement (BCL6) – Positive

• B‐cell lymphoma, unclassifiable, with features intermediate between DLBCL and Burkitt lymphoma (Double‐hit)

• Predicted prognosis ‐ Poor

169

Conclusions• Aggressive B‐cell lymphomas are

clinically, histologically and genetically diverse neoplasia.

• Molecular genetics is necessary to convey important prognostic information relevant for therapy.

• Evaluate all DLBCL with greater than 95% PI and other aggressive B cell lymphomas with the panel of FISH probes.

Double HitLymphoma

DLBCL

BL

MYC BCL2

DHS = 1

DHS = 2Green TM et al., JCO 30:3460-3467: 2012171

MYC/BCL2 immunohistochemistry in DLBCL


• MYC/BCL2 coexpression was reasonably sensitive and specific for MYC/BCL2 double‐hit rearrangements: sensitivity 79%, specificity 81%

• Much better than Ki‐67 index (recent study used 75% as threshold, Maitong‐Kalaw et al., Histopathology 2012;61:1214‐8)

172Johnson NA et al., J Clin Oncol. 2012

Green TM et al., J Clin Oncol. 2012

• Coexpression of MYC and BCL2 identifies a subset of patients with DLBCL (18‐29%) with inferior OS and PFS: “biologic double‐hit”

• Independent of IPI and cell of origin (ABC and GCB)

• Co‐expression of MYC and BCL2 can identify the ABC‐type and IPI‐high DLBCL cases with worse outcomes

• May identify the majority of patients with refractory or resistant DLBCL

173


Impact of novel genetic alterations identified by next generation

sequencing

Aggressive B‐cell lymphomasin 2015

Mutations of ID3 and TCF3 cooperatewith IG‐MYC in BL

ID3 and TCF3mutations in 70% of sporadic BL6/47 (13%) other BCL with MYC‐IG translocation

Cyclin D3 in mutations in 38% sporadic BL .

175

Cooperation between MYC and ID3/TCF3/Cyclin D3 in BL

Schmitz R et al., Nature 2012176

Relative frequency of gene alterations in GCB and ABC DLBCL

Courtesy of Bailey N 2013177

178

Impact of genetic aberrations on pathophysiology of B‐cell NHL

TAKE HOME MESSAGES• True BL has low genetic complexity

• Identification of MYC positive mature lymphoma is important to guide chemo (CVAD vs. R‐CHOP)– mBL signature (t(8;14), low complexity,benefit from CVAD

– Double hit or MYC+ DLBCL (high complexity), less responsive to either

179

Take home messages• Aggressive B‐cell lymphomas are clinically, histologically and genetically diverse neoplasms

• Molecular genetics is necessary to convey important prognostic information relevant for therapy

• Algorithmic approach using immunohistochemistry and FISH analysis will allow subclassification for better risk stratification and treatment

• NGS identifies novel genetic events that may have a role in diagnosis and therapy

180

Answers for Case 3• What are the differential diagnostic considerations?

DLBCL, BL, Double hit lymphoma, MCL, BCL Unc

• What additional stains would you perform to refine the diagnosis?CD20, BCL2, BCL6, MIB1, MUM1

• What additional molecular studies may help refine the diagnosis?FISH for MYC, BCL2 and BCL6 rearrangements

• What is the role of karyotypic analysis, FISH and gene rearrangement in the workup of this class of lymphoma?

Gene rearrangement study would not be helpfulKaryotypic and FISH studies will aid in determining genetic complexity and presence of MYC, BCL2, BCL6 rearrangements

181

Answers for Case 3

What other lymphoid neoplasms demonstrate MYC aberrations?

ALCL, MCL, NK/T cell lymphomasWhat genetic abnormalities have been recently described in the group of aggressive B‐cell lymphomas?

ID3, TCF3, Cyclin D3 mutations in BLCARD11,CD79a,A20 mutations in DLBCL

182

Case 4

• A 53 year old man presented with abdominal discomfort. Physical examination and radiologic studies revealed an enlarged spleen which was removed. The spleen weighed 3.55 kg with dimensions of 25 x 19.5 x11.7 cm.

• A virtual slide of a section of the splenectomyspecimen is provided.

183

Case 4: H&E

184

Case 4• What are the differential diagnostic considerations?• What additional stains would you perform to refine the

diagnosis?• What additional molecular studies may help refine the

diagnosis?• What is the role of karyotypic analysis, FISH and gene

rearrangement in the workup of splenic lymphomas? • What genetic abnormalities have been recently described

in lymphomas that present primarily in the spleen?• How will genetic abnormalities contribute to patient

management such as prognosis, therapy and disease monitoring?

185

Case 4: Differential diagnosis

• Splenic marginal zone lymphoma• Splenic B‐cell lymphoma, unclassifiable• Follicular lymphoma• Mantle cell lymphoma• Chronic lymphocytic leukemia/SLL• Hairy cell leukemia

• T cell lymphoma

186

Case 4: Immunophenotype

CD5 negativeCD43 negativeCyclin D1 negativeCD10 negative

187

Splenic Marginal Zone Lyphoma

• SMZL uncommon indolent lymphoma involving splenic white pulp, blood, and bone marrow

• First-line therapies• splenectomy• anti-B-cell biologicals

• Median survival 10yr

188 Kiel MJ, .. Elenitoba-Johnson KSJ J Exp Med. 2012 Aug27;209(9):1553‐65.

PathoPic

EH&O

Molecular genetics of splenic lymphoma

189

• del7q, +3/+3q [+18, +12] • no recurrent translocation• no known genetic etiology

• del7q, +3/+3q [+18, +12] • no recurrent translocation• no known genetic etiology

Experiment Design: Whole Genome and Targeted Sequencing

190

del7q

del7q

-19

del13q

+8q

SMZL Genome Complexity

191

NOTCH2 Frameshift Mutation

192

Recurrent NOTCH2 Nonsense Mutations

193

Recurrent NOTCH2 Mutations in SMZL

Additional 93 SMZL specimens sequenced in validation cohort.22 additional cases with NOTCH2 mutations identified

Kiel MJ, .. Elenitoba-Johnson KSJ J Exp Med. 2012 Aug27;209(9):1553‐65. 194

Frequency of NOTCH2 Mutations in Various Lymphoma Subtypes and Reactive Lymph Nodes


Recurrently targeted pathways in SMZL

Rossi D et al. J Exp Med 2012;209:1537-1551 196

Decreased Relapse-free Survival in NOTCH2-mutated SMZL


SMZL

SMZL

delta jagged/serrate

γ-secretase

NICD

CSLCoR CSL

NICDCoA

Transcription of target genes:HES, HEY, NF-κB, PPAR

CoR

Notch heterodimer

Notch signaling can be inhibited by gamma secretase inhibitors

DAPT

Cell survival and differentiation

198

ConclusionsConclusions

• NOTCH2 is recurrently mutated in SMZL• Mutations cluster in C-terminus

causing gain-of-function of NOTCH2• NOTCH2 mutations are specific to MZL• NOTCH2 mutations confer an adverse

prognosis

• KLF2 is recurrently mutated in SMZL

199Piva R et al., Leukemia 2014 KLF2

Molecular testing in small B cell lymphomas of the spleen

200

• Splenic marginal zone lymphoma

• Splenic B‐cell lymphoma, unclassifiable

• Nonsplenic MALT• Follicular lymphoma• Mantle cell lymphoma• Chronic lymphocytic leukemia/SLL

• Hairy cell leukemia

Novel alterations in lymphoid neoplasia identified by NGS

201

Recurrently mutated genes in CLL

202

Gene Protein Mutation Mutated cases /

total

Overall frequency

(%)

Frequency in IGHV-

unmutated(%)

Frequency in IGHV-

mutated (%)

NOTCH1 Notch 1 P2515Rfs*4Q2503*F2482Ffs*2

29/2551/2551/255

12.2 20.4 7

MYD88 Myeloid differentiation primary response gene 88

L265P 9/310 2.9 0.8 5.6

XPO1 Exportin 1 E571KE571G

3/1651/165

2.4 4.6 0

KLHL6 Kelch-like 6 F49L/L65PL90FL58P/T64A/Q81P

3/160 1.8 0 4.5

Puente XS et al. Nature, 2011; 475: 101-05.

N‐Ras

B‐Raf

MEK1/2

ERK1/2

A MOLECULAR DIAGNOSTIC APPROACH

BRAF V600E is strongly associated with classic HCL

203

BRAF exon‐15 mutation is sensitive and specific for hairy cell leukemia

204

Cases (n=243)Tumor Entity

analyzed mutated % mutatedHairy Cell Leukemia 48 48 100Splenic Marginal Zone Lymphoma 22 0 0Splenic Lymphoma/Leukemia, Unclassifiable* 16 0 0Chronic Lymphocytic Leukemia 21 0 0Follicular Lymphoma 35 0 0Diffuse Large B-cell Lymphoma 71 0 0Mantle Cell Lymphoma 18 0 0Burkitt Lymphoma 12 0 0

Tiacci E et al. N Engl J Med 2011;364:2305-2315.

N‐Ras

B‐Raf

MEK1/2

ERK1/2

MAP2K1 (MEK) mutations are associated with HCL‐v

205

RAF kinase is a molecular target

206Vakiani and Solit J Pathol 2011: 223:219-229

Improved survival with vemerafanib in melanoma patients with BRAF

mutations

207Chapman PB et al., N Eng J Med 2011: 364:2507-2516

Immunohistochemistry for detection of B‐RAF V600E (clone VE1) mutant protein

208Andrulis M et al., Am J Surg Pathol 2012

B‐RAF V600E (clone VE1) mutant protein IHC for minimal residual disease

209Akarca AU et al., Br J Haematol Jul 2013

Discovery of novel genetic alterations in B‐NHL

210

Traditional NGSBL MYC rearrangements ID3, TCF3, Cyclin D3

DLBCL, GC MYC, BCL2, BCL6 rearrangementsSomatic hypermutations

Chromatin remodelling genes(MLL2, EP300, CREBBP)Epigenetic modifiers (EZH2)TBL1XR1/TP63

DLBCL, activated MYC, BCL2, BCL6 rearrangements CARD11, CD79B, A20,BLIMP1. MYD88Chromatin remodelling genes(MLL2, EP300, CREBBP)

Pediatric DLBCL MYC IRF4 translocations

Primary mediastinal B‐cell lymphoma JAK2 amplifications MHC) class II transactivator CIITA

Hodgkin lymphoma JAK MHC) class II transactivator CIITA

Splenic marginal zone lymphoma del7q, +3/+3q [+18, +12] NOTCH2, NFKB pathway genes, chromatin remodelling genes

Hairy cell leukemia IGH‐CyclinD1 B‐RAF/ MAP2K1

Follicular lymphoma IGH‐BCL2 EZH2, MLL, TBL1XR1/TP63

CLL TrisomyDel 17p, del11q, del13q

NOTCH1, MYD88, SF3B1

Waldenstrom macroglobulinemia MYD88

Answers for Case 4• What are the differential diagnostic considerations?

Splenic marginal zone lymphomaSplenic B‐cell lymphoma, unclassifiableNonsplenic MALTFollicular lymphomaMantle cell lymphomaChronic lymphocytic leukemia/SLLHairy cell leukemia

• What additional stains would you perform to refine the diagnosis?CD43, CD5, Cyclin D1, CD25

• What additional molecular studies may help refine the diagnosis?B‐RAF for HCL, MYD88 for WM

• What is the role of karyotypic analysis, FISH and gene rearrangement in the workup of splenic lymphomas?

FISH for CLL associated genetic alterationsFISH for t(11;14)

211

Answers for Case 4

• What genetic abnormalities have been recently described in lymphomas that present primarily in the spleen?

NOTCH2mutations in 25% KLF2mutations in 40%

• How will genetic abnormalities contribute to patient management such as prognosis, therapy and disease monitoring?

SMZL with NOTCH2mutations may have worse survivalNOTCH2 signaling may represent a therapeutic target

212

Take home messagesTake home messages

• Many subtypes of B‐cell lymphomas can present in the spleen

• SMZL is a B‐cell malignancy that has not been associated with a recurrent genetic abnormality

• NGS studies identified NOTCH2mutations in 25% of SMZL

• Gene mutations in NFKB pathway, chromatin remodeling are also present in SMZL

• Molecular studies may help in subclassification of other B‐cell lymphomas that present in the spleen

213

Case 5

• A 14 year old boy presented with a one year history of diffuse lymphadenopathy of cervical, axillary, abdominal regions. He complained of fevers, weight loss, night sweats.

• An excisional biopsy of the cervical lymph node was performed. A virtual slide of the lymph biopsy is provided.

214

Case 5: Questions• What are the differential diagnostic considerations?• What additional stains would you perform to refine the diagnosis?

• What additional molecular studies may help refine the diagnosis?

• What is the role of karyotypic analysis, FISH and gene rearrangement in the workup of T cell lymphomas?

• What genetic abnormalities have been recently described in T cell lymphomas?


215

Case 5: H&E

216

Differential diagnostic considerations

Hematopoietic • Non‐Hodgkin lymphoma

‐diffuse large B‐cell lymphomas with anaplastic features, ‐anaplastic large cell lymphoma (ALK‐positive and ALK‐negative)Peripheral T cell lymphomas, NOS

• Extramedullary myeloid tumors, • Hodgkin lymphoma • Anaplastic myeloma

Non‐hematopoietic • Melanoma • Carcinoma (anaplastic

variants)

217

Immunophenotype

218

CD30

ALK1

Immunophenotype

Antibody ResultsCD45 Negative

CD20 Negative

CD3 Negative

CD2 Negative

CD4 Positive

CD7 Negative

CD8 Negative

CD30 Positive +++

Perforin Positive

ALK‐1 Positive +++ N/C

Anaplastic large cell lymphoma, ALK+

Diagnosis

219

ALKALK

ALKALK

Y Y YYJAK2/3 PI3K

AKTPDK

STAT3/5 PKCα

NPM

AUTOPHOSPHORYLATION

NPM

Apoptosis Proliferation ??

t(2;5)(p23;q35)

PLC

Signaling Cascades Induced byNPM-ALK

Activation

Localization

Expression

Aberrant

ALK

220

t(2;5)(p23;q35) ~75% N/C 80kd

t(1;2)(q25;p23 ~18% C 104kd

t(2;3)(p23;q21) ~1% C 97kd

t(2;3)(p23;q21) ~1% C

inv(2)(p23q35) ~2% C 96kd

t(2;17)(p23;q23) ~2% C 250ld

t(2;19)(p23;p13.1) - C

t(2;2)(p23;q11-13)?or inv (2)(p23p11-13)?

- C

ALK Kinase

ALK Kinase

ALK Kinase

ALK Kinase

ALK Kinase

ALK Kinase

ALK Kinase

ALK Kinase

NPM

TPM3

TFGL

TFGS

ATIC

CLTC

TPM4

RanBP2

117 680

221 784

193 756

138 701

229 792

1634 2197

221 784

867 1430

Variant ALK Translocation Partner Genes

CARS, Moesin, MYH9, SEC31A221

Diagnosis by FISH Cytogenetics

222

Intracellular localization of ALK expression is dependent on the partner gene

Nuclear and cytoplasmic Cytoplasmic

223

Other tumors that express ALK protein

• Lymphomas

• Soft tissue tumors

• Carcinomas

• Neuroblastomas

• Chromosomal translocations

• Gene amplifications• Kinase activating

mutations• Overexpression

MechanismsNeoplasms

224

ALK 75%

15%

ALCL2p235q35

NPM1

TMP3

TFG

ATIC

CLTCL

~10%

IMT

Diffuse large B-cell lymphoma

TPM4

ALK

ALK

ALK

ALK

ALK

MYH

MSN

ALO17

ALK

ALK

ALK

RanBP2 ALK

ATIC ALK

CLTCL ALK

TPM4 ALK

TMP3 ALK

CARS ALK

SEC31L1 ALK

EML4

KIF5B

ALK

ALK

CLTCL ALK

ALKNPM1

Non-small-cell lung cancer

ALK translocations in human cancer

SQSTM1 ALK

225

ALK is a therapeutic target

Phase 1/2 study of PF-2341066, oral small molecule inhibitor of ALK and C-MET in children with relapsed/refractory solid tumors and anaplastic large cell lymphoma

ADVL0912

Children’s Oncology Group

Butrynski JE et al: N Engl J Med 2010;363:1727-33Kwak EL, et al: N Engl J Med 2010;363:1693-703

Phase II study of crizotinib in children with newly diagnosed anaplastic large cell lymphoma

ANDHL12P1Children’s Oncology Group

226YP Mosse , MS Lim et al Lancet Oncology 2013 May;14(6):472‐80

227

580

4126

77 8 3 60

500

1000

1500

2000

2500

3000

3500

4000

4500

0 1 2 3 4

Copies NPM

‐ALK/10 00

0 AB

L

Months post treatment

Molecular monitoring of NPM-ALK transcript in bone marrowand peripheral blood samples of ALCL patients before and after Crizotinib

38

59

23

913 13

6 71 3 3 1

0

10

20

30

40

50

60

70

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Copies NPM

‐ALK/10 00

0 AB

L

Months post treatment

YP Mosse , MS Lim et alLancet Oncology 2013 May;14(6):472-80

Update on genomic studies of mature T cell lymphoma/leukemia

228

• JAK-STAT mutations in 76% of T-PLL

• IL2RG implicated in human cancer for the first time

• Inhibition of JAK or pSTAT5 may represent a therapeutic strategy for T-PLL patients

229

Integrated genomic analysis of T-PLL identifies novel highly recurrent activating mutations

Kiel M et al., Blood. 2014 Aug 28;124(9):1460‐72

JAK1 JAK3JAK3

STAT5B

IL2

IL2R

IL2R

IL2R

Transcriptional activation

230BLOOD, 11 Dec 2014: 124: 3768

TYK2 gene fusions in cutaneous CD30+ LPDs

CD30

TYK2TYK2

NPM1‐TYK2

Recurrent structural abnormalities in T‐cell neoplasia

231

Disease entity Structural Abnormality Frequency

ALK+ ALCL NPM1‐ALK

Various‐ALK

84%

16%

T‐PLL TRA‐TCL1A

TRA‐MTCP1

70‐80%

10%

HSTL i(7q)(q10) >80%

EATL 9q gains

16q12.1 loss

~70%

~30%

ALK‐ ALCL, c‐ALCL IRF4/DUSP22 translocations

TYK2/ NPM1‐TYK2

25%

17%/4%

PTCL P53‐related genes 6%

F‐PTCL ITK‐SYK 18‐38%

AITL ITK‐SYK rare

Recurrent somatic gene mutations in T‐cell neoplasia

232

Leukemias

Lymphomas

Gene Disease entity FrequencyJAK1/JAK3 T‐PLL

ENKTCL

40%

20‐35%

STAT3 T‐LGLLGD HSTCLCLPD‐NK

27‐40%9%30%

STAT5B T‐PLLGD HSTCLT‐LGLL

36%36%rare

RHOA AITL; PTCL, NOS 67%; 18%

FYN AITL; PTCL, NOS 3% rare

TET2 AITL

F‐PTCL

33‐47%

58%

IDH2 AITL ~25%

DNMT3A AITL; PTCL, NOS 11% overall

Molecular studies will impact therapeutic decisionsin T‐cell neoplasia

Lymphoma Genetic features Therapeutic relevance

ALCL ALK ALK inhibitorcALCL TYK2 TYK2 inhibitorAITL ITK/SYK SYK inhibitorT‐PLL/ENKTCL JAK3 JAK3 inhibitorT‐PLL STAT5B STAT5 inhibitorT‐LGLL, NK‐LPD STAT3 STAT3 inhibitor

AITL TET2, IDH2, DNMT3A Epigenetic modulators

PTCL, NOS DNMT3A Epigenetic modulators

F‐PTCL TET2 Epigenetic modulators

233

Genomic Classification of Mature T‐cell Lymphoma/Leukemia

Cutaneous

Primary Cutaneous CD30+ T‐cell Disorders

Mycosis Fungoides

Extranodal

NK/TCL Nasal Type Adult T‐cell Leukemia/Lymphoma

T‐cell Large Granular

Lymphocytic Leukemia

Subcutaneous Panniculitis‐like TCL

Leukemic

Enteropathy‐associated TCL

Hepatosplenic TCL

Aggressive NK‐Cell Leukemia

T‐PLL

Mature T‐/NK‐cell

Primary Cutaneous Gamma/Delta TCL

Sézary Syndrome

PTCL,NOS

Nodal

AITLAnaplastic Large Cell Lymphoma (ALK +)

Anaplastic Large Cell Lymphoma (ALK‐)

STAT3RHOA/IDH2/TET2/DNMT3A

TYK2

?????

ALK

IRF4/DUSP

JAK3

JAK/STAT5B

STAT3

234

TP53/ARID1A/MLL

Case 5: Answers• What are the differential diagnostic considerations?

Other ALK+ neoplasms, ALCL ALK neg, • What additional stains would you perform to refine the diagnosis?

Pan T cell antigens • What additional molecular studies may help refine the diagnosis?

ALK FISH, TCR gene rearrangement• What is the role of karyotypic analysis, FISH and gene

rearrangement in the workup of T cell lymphomas?TCR gene rearrangement may be necessary for “null cell phenotype”, FISH may be necessary in cases where ALK staining is weak

235

Case 5: Answers

• What genetic abnormalities have been recently described in T cell lymphomas?

TYK2 fusions in cALCLJAK/STAT5B mutations in T‐PLLIDH1/2, TET2, RhoAmutations in AITL


Molecular methods for disease monitoringDetection of ALK mutations for monitoring disease response to therapy

236

SUMMARY AND CONCLUSIONS

237

IMPLICATIONS OF MOLECULAR GENETICS FOR HEMATOPATHOLOGY?

238

Translating New Genetic Findings into Clinical Practice

• Genomic profiling technologies have identified an explosion of genes associated with various hematolymphoid malignancies

• More clinical studies are needed to sort out which genes are most important prognosticallyand therapeutically.

239

Molecular studies will impact therapeutic decisions in lymphoma

Lymphoma Genetic features Therapeutic relevanceBurkitt MYC/IGH CVAD vs RCHOPDouble Hit DLBCL MYC/IGH; BCL2/IGH Not responsive

BCL6/IGHGastric MALT API2/MALT1 MALT1 inhibitorsHairy cell leukemia B‐RAF mutation B‐RAF inhibitorsCLL/SLL IgH SHM No need for

aggressive TxCLL/SLL; SMZL NOTCH mutation Gamma secretase inhibitorFCL/DLBCL EZH2, MLL, p300 Demethylating agentsALCL ALK Tyrosine kinase inhibitorT cell lymphoma ITK/SYK SYK inhibitor

240

Target Year Target Discovered

Disease(s) and Proportions

Estimated Total # Pts Annually (US)

Drug(s) Clinical Outcomes

Outcomes from Conventional Chemotherapy

Year Mutation-Targeted Treatment Documented

BCR-ABL 1960 CML (100%) 5,000ImatinibDasatinibNilotinib

RR 90%5y PFS 80%5y OS 90%

RR 35%5y OS 70% 2001

EGFR 1978EGFR mutated NSCLC (10% of NSCLC)

17,000 ErlotinibGefitinib

RR 75%Median PFS 11 mosMedian OS 31 mos


2004

KIT 1998 GIST 6,000 Imatinib


RR 5%Median OS 20 mos 2002

BRAF 2002 50% of melanoma100% HCL 34,000 PLX4032

RR 77%Median PFS 7 mosOS not yet determined

RR 10-20%PFS 1.5 mosOS 8 mos.

2010

ALK 2007EML4-ALK NSCLC (5% of NSCLC)ALK+ ALCL

8,500 Crizotinib

RR 55%6 month PFS 70%OS not yet determined

RR 25%Median PFS 4-6 mosMedian OS 12 mos

2010

Changing Pace of Target Discovery to Therapy

Gerber DE and Minna JD Cancer Cell 2010241

242

Multiplexed targeted platforms

243

Impact on WHO Classification?

WHO 2008 Classification

Diagnosis Prognosis Therapy

CONCLUSIONSCONCLUSIONS• Molecular techniques are routinely being employed to

provide adjunctive results critical to patient management

• Molecular techniques also provide opportunities for improved diagnosis, early disease detection and prognosis of hematopoietic diseases

• Molecular techniques e.g. ABL1 kinase sequencing now offer opportunities for therapeutic refinement and adaptation to diseases specific to each patient (personalized medicine)

244

• New genomic tools have led to discovery of novel mutations in hematopoietic neoplasms.

• Certain genetic abnormalities previously underappreciated as important lesions in hematopoietic malignancies.

• As costs of sequencing continue to fall, additional novel mutations likely to be identified at a rapid rate.

245

CONCLUSIONSCONCLUSIONS

1980 1985 19901988 1994 2001 2008 …

Morphologic evaluation

Cytogenetics

Introduction of immunopathology, flow cytometry immunohistochemistry

Rapid growth in molecular genetic application

Gleevecapproved by FDA

New therapies

•Combination of small molecules

Multi-panel detection of structural alterations/mutations with prognostic and therapeutic implications

2015

Evolution of Molecular Diagnostics in Hematopathology

PCR FFPE tissues REAL classification

WHO classification

246

NEED FOR INTEGRATED HEMATOPATHOLOGY REPORTS

247

1) What disease does the patient have? (diagnosis)

1) How much of the disease is there? (residual disease)

2) What drug will the disease respond to? (therapy)

3) Who needs treatment? (prognosis)

4) What dose? (pharmacogenomics, dynamics)

Mass spectrometry

Integrated evaluation of molecular abnormalities in hematologic disease

New insights into pathogenesis

Altered gene expressionAltered gene expression

Gene rearrangementsGene rearrangements Gene mutationsGene mutationsChromosomal translocationsChromosomal translocations

Early Detection Diagnostics Prognostics New Therapeutics 248

Next Generation Pathologist

SNP arrayaCGH NGS

EpigenomeEpigenome

Thank you for participating!Please complete the course evaluation before you leave.

249

Documents

Impact of Personalized Medicine for the Practice of Hematopathology