Minimale residuele ziekte (MRD) detectie in acute leukemie: … · 2014. 2. 6. · • Prognostic relevance of patients undergoing stem cell transplantation ... Cross-lineageantigen

4-2-2014

1

© 2010 Universitair Ziekenhuis Gent

Minimale residuele ziekte (MRD)

detectie in acute leukemie:AML/ALL

Apr. Biol. Barbara Denys

Lab Klinische Biologie

UZ Gent

VAKB

3 februari 20142© 2010 Universitair Ziekenhuis Gent

ALLALL MM MM CLLCLL LymphomasLymphomas

Hematopoietic

stem cellNeutrophils

Eosinophils

Basophils

Monocytes

Platelets

Red cells

Myeloid

progenitor

Myeloproliferative disordersMyeloproliferative disordersAMLAML

Lymphoid

progenitor T-lymphocytes

Plasma

cells

B-lymphocytes

nanaïïveve

3© 2010 Universitair Ziekenhuis Gent

Overview presentation

• Introduction• Definition of MRD – goal of MRD• MRD techniques• Clinical value of MRD in AML/ALL

�Prognostic significance?

�MRD directed therapy?

• Future perspectives


Incidence of AML and ALL


Overall Survival in AML patients

- Primary induction failure:

- Young pt: 20-30%

- Eldery pt: 40-50%

- Relapse: high %, afh RF

- Early relapse: low rate of

CR2 (20%)

- Survival alfter relapse:

10%


Survival ALL patients

4-2-2014

2


0 1 2 3 4

freq

uenc

yof

leuk

emic

cells

follow-up in years

“ cure”

100

0

10

1

0.1

0.01

0.001

(%)

relapse

Information on the treatment effectiveness

MRD detection

MRD


Factors involved in treatment effectiveness


• Identification/detection of small subsets residual leukemic cells

• Kinetics: tumor load reduction during and after induction treatment provides crucial information about the respons to treatment

• Prognosis:• Predictor of outcome of patients with leukemia• Prediction of relapse• MRD-based stratification: identification of low-risk

(therapy reduction) and high-risk (therapy intensification) patients

• Prognostic relevance of patients undergoing stem cell transplantation

GOAL of MRD detection


0 1 2 3 4

freq

uenc

yof

leuk

emic

cells

follow-up in years

“ cure”

100

0

10

1

0.1

0.01

0.001

(%)

relapse

Information on the treatment effectiveness

MRD detection with sensitive techniques

detection limit ofPCR techniques

detection limit ofcytomorphological

techniques

detection limit ofimmunophenotyping


Techniques for MRD


Identification of Leukemia Associated Immuno-Phenotype or LAIPs,

based on immunophenotypic aberrancies

MRD by flow cytometry

1. Cross-lineage antigen expressione.g. AML with CD2+ / CD7+ / CD56+

2. Asynchronous antigen expression- different maturation: e.g. CD34+/CD117+ with CD14+/CD15+ or

CD34+/CD19+/sIg+

- missing marker: CD33-/HLADR-/CD13- on myeloblasts

3. Over- and under-expression of markers++: HLA-DR/CD117/CD13/CD33/CD34/10

- - : CD34/CD117/CD45/24

4. Ectopic antigen expressionNG+ or TdT+/CD3+/CD5+

4-2-2014

3


Examples of different LAIPs

R1

B-ALL: under-expressionB-ALL: over-expression

T-ALL: ectopic expressionAML : cross-lineage expression

normal

normal


Immunophenotypic signature of leukemic cells

Advantages:

• Sensitivity between 10E-4 and 10E-5 with 6-color flow cytometry (multicolor analysis)

• Widely accessible

• Rapid delivery of results• Simple; one technique

• Applicable to almost all patients

• Information about cell viability and status of normal hematopoietic cells in the sample studies

• Single-cell level, allowing recognition and characterisation of small subpopulations

Immunophenotypic detection of MRD


• Heterogeneity of blast cells, especially in AML (subsets in ~75% of AML patients); preferably all subpopulations should be monitored

Major limitations of MRD by flow cytometry (1)


• Heterogeneity of blast cells, especially in AML (subsets in ~75% of AML patients); preferably all subpopulations should be monitored

• Lack of ‘a priori’ (e.g. at diagnosis) of the sensitivity for each individual patient later on during FU of the disease

• Expertise and knowlegde required for LAIP recognition

• Lower sensitivity (10E-3 – 10E-4) than PCR analysis using 3-4 color flow cytometry

• Difficult to standardize (inter-laboratory)

• LAIP• Absence of a clear LAIP at diagnosis

• Low frequency in normal marrow (regenerative marrow)



Background of LAIP in normal bone marrow

Feller et al, Blood Cancer Journal 2013

Aberant marker expression on

normal BM: % of primitive marker

compartment (CD34 or CD117)


• Immunophenotypic shifts

� Most of these changes involve Ag of cell maturation (CD10,

CD22, TdT, CD34, CD13, CD33)

� Abnormal/infrequent antigenic markers are often stable

� Preferably ≥2 patient-specific labelings should be used per

patient; in vast majority of patients at least one MRD labeling

remains informative


4-2-2014

4


• Enormous increase in number of data by merging and calculations

• Automated method for analysis of flow cytometry immunophenotypic data

• Reference picture will facilitate MRD analysis

• Reducing expert-based data-analysis

Flow cytometry: Improvement of data analysis

• Based on visualization of multiple bi-dimensional plots• Operator’s selection of population of interest (subjective)• Depending on the expertise of the operator

Conventional methods of manual data analysis


New software tools - Maturation


Molecular techniques for MRD detection

PCR analysis of genetic abnormalities:

� Fusion genes

� Ig/TCR gene rearrangements


Fusin genes in B-ALL

Pui et al., NEJM 350:1535-1548 (2004)


Fusion genes in AMLAdult

Pediatric

24© 2010 Universitair Ziekenhuis Gent Grimwade D, Morzek K; Hematol Oncol Clin N Am 2011

Importance of gene fusions = PROGNOSIS

4-2-2014

5


Overview quantitative qPCR UZ Gent

Target

(transl/mut/over

expr)

Gene Fusions Incidence Prognosis

t(9;22) (q34;q11) ABL/BCR: p190 en

p210

Pre B-cel ALL: 25-30%

volwassenen; 2-5% kinderen

Slechte prognose

t(8;21) (q22;q22) ETO/AML1 AML: 5-12 %, AML M2: 30 % Goede prognose

t(4;11) (q21;q23) V/MLL

11q23

abnormaliteiten

5% pre B-cel ALL

Meest frequente 11q23

abnormaliteit bij pre B-ALL

Slechte prognose

t(9;11) (p22;q23) MLL-AF9 Precursor B-cel ALL

11q23: 5-6 % van AML (M4-5)

Intermediaire

prognose

t(12;21)(q12;q22) TEL/AML-1 25% pre B-ALL bij kinderen Goede prognose

inv(16) (p13;q22) MYH11/CBFbeta 8-12 % van AML, zeer hoge

associatie met AML M4 eos

Goede prognose

t(1;19) (q23;p13) E2A-PBX1 3-5% kinderen - 3%

volwassen ALL

prognose

controversieel

t(15;17)(q22;q21) PML/RARalfa 5-8 % van AML, associatie

met AML M3: > 95%

Goede prognose26© 2010 Universitair Ziekenhuis Gent

Overview qualitative RT-PCR UZ Gent

Target

(transl/mut/overexpr)


t(6;9) (p23;q34) DEK/CAN 1% AML Slechte prognose

t(1;22)(p13;q13) OTT-MAL AML7 (children) Slechte prognose

Microdeletie 1p32 SIL-TAL 5-25% T-ALL ? controversieel

Target

(transl/mut/over

expr)


NPM1 mutatie / 25-35% AML (~50% CN-

AML)

Good

WT1-overexpressie / 80-90% /

Overview quantitative qPCR UZ Gent


PCR analysis of fusion transcripts

Advantages

�High sensitivity

�Stable

�Disease-specific

�Fast

�Relatively easy and

cheap

�Efforts for

standardization (EAC, Leukemia 2003)

Disadvantages

�Applicable to only

minority of patients

•B-ALL: 40-45%

•T-ALL: 15-35%

•AML: ~30%

�Instability of RNA

�Inter-laboratory

variation


Fusion gene expression with a quantitative ‘real-time’ RT-PCR

EAC protocol

LEADING ARTICLE:

“Standardization and quality control studies of ‘real-time’ quantitative

reverse transcriptase polymerase chain reaction of fusion gene

transcripts for residual disease detection in leukemia” – A Europe

Against Cancer Program

J Gabert, E Beillard, VHJ van der Velden, W Bi, D Grimwade, N Pallisgaard, G Barbany, G

Cazzaniga, JM Cayuela, H Cave, F Pane, JLE Aerts, D De Micheli, X Thirion, V Pradel, M

Gonza´lez, S Viehmann,M Malec, G Saglio and JJM van Dongen

Leukemia 2003

Kwantificatie

Standaardisatie


MRD detection with fusion genes

• No consensus regarding cut-off levels or most predictive threshold

• Not possible to accurately calculate MRD levels, because the number of transcripts produced by each

cell is unknown

• PB versus BM MRD detection?• Timing of PCR assessment? When?


PB versus BM MRD detection of fusion transcripts?

Question of sensitivity? Does the presence of circulating blasts

reflect the persistence of malignant cells in the BM?

Literature:

� qPCR to monitor CBF-AML (AML-ETO en inv16): comparable

sensitivity (Buccisano et al, Blood 2012);

During treatment, quantitative differences -> better BM

After end of treatment, FU on PB possible (10%-15% discrepancies)

� PML-RARA: during therapy BM = PB but after consolidation

BM 1.5 log more sensitive

� WT1: less sensitive ~ background expression in BM (normal HP

stem cells & myeloïd progenitors) -> PB is more specific (low

background in PB)

4-2-2014

6


Ig/TCR gene rearrangements for MRD analysis

• Rearrangement of TCR en Ig genes : ‘fingerprint-like’ or unique rearranged DNA

� Recombination of Variable (V), Diversity (D) en Joining (J)

segments

� Random insertion of nucleotides

� Hypervariable (CDR3) regions with junctional diversity

Van der Velden et al. Leukemia 2003 32© 2010 Universitair Ziekenhuis Gent

MRD assesment based on ASO-qPCR

• Target identification

Kindly provided by Marleen Bakkus, UZ Brussel


MRD assesment based on ASO-qPCR

• ASO design / primer-probe development


PCR analysis Ig/TCR rearrangements

Advantages

�High degree of

standardization (ESG-MRD-ALL, since 2010 EURO-MRD):

design, interpretation and

guidelines

�High sensitivity

�Applicable for most ALL

patients (90-95%)

�Stability of DNA

Disadvantages

�Time consuming

�Expensive

�Need for preferably 2

PCR targets (clonal

evolution)

�Extensive expertise is

needed


MRD by PCR (fusion genes) vs FLOW in AML

• Study of Inaba et al (JCO 2012)

• 203 AML (children)

• 1215 compared samples• MRD of fusion genes

vs FLOW 4-colors

• Poor correlation• FLOW-/PCR-: 99%

• FLOW+/PCR+: 19%

Inaba et al, JCO 201236© 2010 Universitair Ziekenhuis Gent

Flow cytometry vs PCR in ALL

cut-off 0.1%N FCM-/PCR- FCM+/PCR+ FCM+/PCR- FCM-/PCR+ concordance

day 15 174 24,1% 69,0% 2,9% 4,0% 93,1%day 33 187 81,3% 13,4% 2,7% 2,7% 94,7%day 78 198 96,0% 3,0% 0,5% 0,5% 99,0%All 559 68,7% 27,0% 2,0% 2,3% 95,7%

cut-off: 0.01% 1

N FCM-/PCR- FCM+/PCR+ FCM+/PCR- FCM-/PCR+ concordanceday 15 163 3,7% 93,9% 0,6% 1,8% 97,5%day 33 158 60,8% 32,9% 2,5% 3,8% 93,7%day 78 185 91,4% 4,9% 0,5% 3,2% 96,2%All 506 53,6% 42,3% 1,2% 3,0% 95,8%

no cut-offN FCM-/PCR- FCM+/PCR+ FCM+/PCR- FCM-/PCR+ concordance

day 15 174 1,1% 94,8% 0,0% 4,0% 96,0day 33 187 33,7% 39,6% 0,5% 26,2% 73,3day 78 198 70,7% 8,1% 0,0% 21,2% 78,8All 559 36,7% 45,6% 0,2% 17,5% 82,3

Literature for ALL: 3/4 color FCM MRD vs PCR-based MRD

-> 70-95% concordance (cut-off 0.01%)

Dutch DCOG-ALL10 protocol for childhood ALL: excellent concordance,

over 95% with 6-color FCM and with cut-off 0.01%

Denys et al,

Leukemia 2013

4-2-2014

7


Clinical significance of MRD?

• Does MRD has clinical significance?• Does it predict relapse?

• Has it an independent prognostic factor?

• What time-points of analysis during therapy are informative?

• What cut-off levels (threshold for positivity) are informative?

• Is there any difference beween AML and ALL? Between adult AML and childhood AML?

• Clinical intervention quided by MRD results?38© 2010 Universitair Ziekenhuis Gent

Prognostic value of MRD in ALL


MRD prognosis in childhood ALL

“MRD after induction is most important independent

prognostic factor!”

Cavé et al, NEJM 199840© 2010 Universitair Ziekenhuis Gent

Treatment scheme childhood ALL in frontline


Treatment of MRD post-transplant Phi+ ALL

• Detection of MRD post-transplant = 90% probability of relapse

• Start imatinib in MRD setting

• Post-transplant imatinib group: significant better OS (86.7% vs 34.3%)

Wassmann et al, Blood 2005

Chen et al, J Hematol Oncol 2012


Prognostic value of MRD in AML

4-2-2014

8


Prognostic significance of MRD in AML (adult)

FLOW FLOW44© 2010 Universitair Ziekenhuis Gent

Prognostic significance in AML (adult)

FLOW FLOW


Post-ind1: Mean 0.040% (0.01-16%)

Post ind2: Mean 0.023% (0.01-21%)

Post Cons (chemo, alloSCT of autoSCT)Mean 0.021% (0.01-9.6%)

� Cut-off = 0.1%

�MRD (FCM)= independent

prognostic factor

�Retrospective multivariate

analysis:

MRD+ after induction (2 cycles) =

significant higher relapse rate (RR)


HOVON SAKK/AML 42A; Terwijn et al, JCO 2013 46© 2010 Universitair Ziekenhuis Gent

MRD inv(16)/t(8;21) in AML

Importance of PCR negativity combining consolidation samples points

and first 3 months of follow-up (Corbacioglu et al, JCO 2010)

Corbacioglu A et al. JCO 2010;28:3724-3729

< 2 PCR-neg samples

2 PCR-neg samples

P=0.001

2 PCR-neg samples

< 2 PCR-neg samples

P=0.01During

consolidation +

first 3 months FU

CBFB-MYH11 AML (n = 53)


MRD with NPM1 mutation-specific qPCR

• Nucleophosmin mutations; good prognosis

• High frequency: 25-30% of AML patiënt (~50% CN-AML)

• Quantitative qPCR (80-90%

type A mutation); stable

marker; high expression

at diagnosis

• CIR according to MRD status

after induction (A en B) and

after completion of therapy

(C en D)

Krönke J et al. JCO 2011;29:2709-271648© 2010 Universitair Ziekenhuis Gent

Prognostic significance of MRD in AML (childhood)

FLOW FLOW

4-2-2014

9


When is MRD assessment clinical relevant?

= Time point with power to provide most informative prognostic

indicator!!!

Optimal sampling interval varies with molecular subgroups

e.g. NPM1: highest relapse risk with high NPM1mut transcript levels after

induction and consolidation / WT overexpression: early monitoring (2 log

reduction after induction) / inv(16): after consolidation

Molecular MRD levels measured at delayed rather than early time

points (superior prognostic relevance) flow cytometry


Conclusion MRD in AML

• MRD is a strong and independent prognostic factor in AML (both childhood and adult)


Overview MRD studies in AML by MPFC


Conclusion MRD in AML

• MRD is a strong and independent prognostic factor in AML (both childhood and adult)

• Relevant cut-off levels vary between 0.01% and 0.5%

• Relevant cut-off levels and time points are protocol dependent

• MRD by flow cytometry: early time points vs molecular: after completion of therapy


Clinical intervention quided by MRD results in AML?


MRD-directed therapy in APL

MRD-directed therapy significant reduction in relapse rate

FIG ...

MRD monitoring = standard inAPL (ELN guidelines – Sanz et al, BLood 2009)

Grimwade et al, Current Opinion

in Oncology 2010

•Therapeutic goal = PCR negativity after consolidation (positivity

results in relapse)

• MRD monitoring:

4-2-2014

10


Clinical intervention guided by MRD for other

molecular targets

• No guidelines• Lack of consistency in the scheduling of MRD

assessment at any given time point in AML

• Only apllicable in a minority of patients• However, real life…


0,00001

0,0001

0,001

0,01

0,1

1

10

12

-11

-20

10

03

-12

-20

10

29

-12

-20

10

07

-02

-20

11

14

-02

-20

11

15

-03

-20

11

24

-05

-20

11

02

-09

-20

11

14

-10

-20

11

17

-11

-20

11

02

-01

-20

12

CB

Fb-M

YH

11

0

10

20

30

40

50

60

70

80

90

100

% b

last

en

CBFB-MYH11 qPCR

% blasten

inductie consolidatie

5Negatief

Example of MRD monitoring in AML patient with CBFB-MYH11


0,00001

0,0001

0,001

0,01

0,1

1

10

12

-11

-20

10

03

-12

-20

10

29

-12

-20

10

07

-02

-20

11

14

-02

-20

11

15

-03

-20

11

24

-05

-20

11

02

-09

-20

11

14

-10

-20

11

17

-11

-20

11

02

-01

-20

12

21

-03

-20

12

11

-05

-20

12

23

-08

-20

12

CB

Fb-M

YH

11

0

10

20

30

40

50

60

70

80

90

100

% b

last

en

CBFB-MYH11 qPCR

% blasten

SCT

5

Example of MRD monitoring in AML patient with CBFB-MYH11

Negatief


Post-ind1: Mean 0.040% (0.01-16%)

Post ind2: Mean 0.023% (0.01-21%)

Post Cons (chemo, alloSCT of autoSCT)Mean 0.021% (0.01-9.6%)

� Cut-off = 0.1%

�MRD (FCM)= independent prognostic

factor

�Retrospective multivariate analysis:

MRD+ after induction (2 cycles) =

significant higher relapse rate (RR)

Prospective patient –tailored risk-based

therapy (clinical decision making):

After induction!

IR with MRD+ after indcution = HR


HOVON SAKK/AML 42A; Terwijn et al, JCO 2013


MRD-directed therapy in childhood AML

Results of the AML02 Multicenter Trial -> FIRST SUDY

MRD used to guide treatment in non-APL AML!

– MRD used to intensify therapy during induction

– MRD to identify patients eligible for HSCT

1. Reduction of refractory

disease (4.6%)

2. Improved OS

Rubnitz etl al, Lancet Oncol 201060© 2010 Universitair Ziekenhuis Gent

Response evaluation on day 22 after first course and second course

with flow cytometry

Inclusion of results in risk stratification!

NOPHO DBH AML2012 (childhood)

4-2-2014

11


Risk groups NOPHO DBH AML2012

FLOW

MRD


Integration of MRD in risk stratification

• Most important risk factors at diagnosis: cytogenetics and molecular abnormalities

• Post-treatment risk indicators = response-related factors -> MRD = treatment-related therapy stratification


Pre-treatment risk factors in AML

ELN paper, Dohner et al, Blood 201064© 2010 Universitair Ziekenhuis Gent

Outcome: combinatie (cyto)genetica en MRD?

Subgroup analysis of RFS and CIR of 143 AML patients stratified according to pretreatment karyotype or FLT3 status and levels of MRD after consolidation.

Buccisano F et al. Blood 2012;119:332-341


Risk assessment combining pretreatment and post-treatment prognosticators.

Buccisano F et al. Blood 2012;119:332-341©2012 by American Society of Hematology


Future

• Risk assessment combining pretreatment and post-treatment prognosticators (MRD included) in AML.

• Improving MRD detection by flow cytometry: focus on stem cells

• Standardisation of techniques (flow and PCR)• Improvements necessary in interpretation and data

analysis (new software tools)

• Next generation sequencing

4-2-2014

12


EINDE

Documents

Minimale residuele ziekte (MRD) detectie in acute leukemie: … · 2014. 2. 6. · • Prognostic relevance of patients undergoing stem cell transplantation ... Cross-lineageantigen