The BIOLOGY of HIGH RISK CLL Gianluca Gaidano, M.D., Ph.D. Division - - PowerPoint PPT Presentation

The BIOLOGY of HIGH RISK CLL

Division of Hematology Department of Translational Medicine University of Eastern Piedmont Novara-Italy

Gianluca Gaidano, M.D., Ph.D.

Disclosures

Roche (Advisory Board) Janssen (Advisory Board) Gilead (Advisory Board) Amgen (Advisory board; research support) Morphosys (Advisory Board) Abbvie (Advisory Board)

Outline

• The genetic landscape of CLL
• Molecular prognosticators of CLL
• Molecular predictors of CLL
• Richter syndrome biomarkers
• Novel biomarkers

Pathogenesis of CLL

Initiation Progression Chemorefractoriness Transformation Microenvironment Interactions Trasforming Lesion Secondary Lesion Predisposition Promotion/Accumulation

Polygenic IRF4 IRF8 MYC Other del13q +12 MYD88 TP53 NOTCH1 SF3B1 BIRC3 ATM MYC CDKN2A Signaling pathways BCR NF-kB TLR CD38 VLA-4 integrins NOTCH CXCR4

CLL is genetically heterogeneous and lacks disease defining genetic lesions

The wordcloud shows the genes that are reported as mutated in CLL by the v77 of the Catalogue of Somatic Mutations in Cancer (COSMIC). The size of the font is proportional to the mutation frequency

• One of the tumor with the lowest background mutation load (0.6 per Mb)
• No unifying gene mutations
• TP53, NOTCH1, SF3B1, ATM mutated in >5% CLL

NF-kB

BIRC3 TRAF3 NFKBIE

NOTCH

TLR

MYD88

Apoptosis

BCL2 del13q (miR15/16) DNA damage response

P

TP53

P

TP53

P

Cell cycle BCR

NOTCH1 FBXW7 SPEN TP53 ATM SF3B RPS15 POT1 CDKN2A

MAPK

RAS BRAF MAP2K1 Moia et al, 2018

Outline

• The genetic landscape of CLL
• Molecular prognosticators of CLL
• Molecular predictors of CLL
• Richter syndrome biomarkers
• Novel biomarkers

Genetic-based models of CLL prognosis

Döhner H, et al. N Engl J Med. 2000 Rossi et al, Blood 2013

(N=325)

Survival

Months 13q-Deletion Trisomy 12

2 4 6 8 10 12 14 0.0 0.2 0.4 0.6 0.8 1.0 Years from diagnosis Cumulative probability of OS (%)

Matched general population

Cytogenetic model Mutational- cytogenetic model

N del13q 26% Normal/+12 40% NOTCH1 M/SF3B1 M/del11q 17% TP53 DIS/BIRC3 DIS 17%

CLL-IPI

Variable Adverse factor Coeff. HR TP53 (17p) deleted and/or mutated 1.442 4.2 Grading 4 Prognostic Score 0 – 10 IGHV status Unmutated 0.941 2.6 B2M, mg/L > 3.5 0.665 2.0 Clinical stage Binet B/C or Rai I-IV 0.499 1.6 Age > 65 years 0.555 1.7 2 1 2 1

Risk group Score Patients N (%) 5-year OS, % Very High 7 – 10 62 (5) 23.3 3.6 High 4 – 6 326 (27) 63.6 1.9 Intermediate 2 – 3 464 (39) 79.4 3.5 Low 0 – 1 340 (29) 93.2

International CLL-IPI working group. Lancet Oncol 2016

Time (months) Overall survival

Low Intermediate High Very high

Overall survival (all patients)

Outline

• The genetic landscape of CLL
• Molecular prognosticators of CLL
• Molecular predictors of CLL
• Richter syndrome biomarkers
• Novel biomarkers

Predictive biomarkers Prognostic biomarkers Treatment tailoring Patient counseling Frequency of follow-up Identify those apropriate for early intervention trials Clinical applications of predictive and prognostic biomarkers in CLL

Antigen-experienced, memory B-cells Naive B-cells Germinal Center B-cells Ag

T-cell dependent affinity maturation

M-CLL MBL

• Genetic lesions
• BCR stimulation
• Microenvironmental interactions

Ag

T-cell independent immune response (no somatic hypermutation)

Antigen-experienced B-cells

U-CLL MBL

IGHV mutated patients gain the greatest benefit from FCR

a p-value vs. matched general population

FCR: fludarabine, cyclophosphamide, rituximab; MDACC: MD Anderson Cancer Center; OS: overall survival; PFS: progression-free survival

• 1. Thompson PA, et al. Blood 2016; 127:303–309. 2. Fischer K, et al.

Blood 2016; 127:208–215. 3. Rossi D et al. Blood 2015; 126 1921–1924.

2 4 6 8 10 12 14 16

p<0.0001

25 50 75 100 PFS (%) Time (years)

PFS (%)

Time (years) PFS (%) Time (years) 25 50 75 100 10 2 4 6 8 OS (%) Time (years) 25 50 75 100 10 2 4 6 8

IGHV mutated IGHV unmutated Matched general population IGHV mutated, FCR IGHV mutated, FC

p<0.001 p<0.0001a p<0.0001a

High-risk (del[17p]) Intermediate-risk (IGHV unmutated and/or del[11q]) Low-risk (IGHV mutated)

p=0.227a

MDACC Phase II study (N=300)1 CLL8 study vs. FC (N=817)2 Italian retrospective analysis (N=404)3

100 25 50 75 2 4 6 8

IGHV unmutated, FC IGHV unmutated, FCR

Hallek et al. Blood. 2008; Oscier et al. Br J Haematol. 2013 Zelenetz et al J Natl Cancer Inst 2015 Eichhorts et al, Ann Oncol 2015.

Guideline recommendations for IGHV analysis in clinical practice Recommendation When iwCLL Not generally indicated

• BCSH

Not recommended

• NCCN

Not generally indicated

• ESMO

Desirable Before treatment

PFS (%) Time (months) 2 4 6 8 10 12 14 16 18 20 22 24 26 100 80 60 40 20 PFS (%) Time (months) 20 40 60 80 100 6 12 18 24 30 36 42 48 54 60 66 72 IGHV mutated (n=16) IGHV unmutated (n=79)

Analysis of ibrutinib-treated patients with R/R CLL/SLL from Phase Ib/II Study PCYC-1102/1103

Sharman JP, et al. ASH 2014, #330; O’Brien S, et al. ASH 2016, #233

Analysis of idelalisib-treated patients with R/R CLL from Phase III Study 116/117

IGHV unmutated (n=91) IGHV mutated (n=19)

BCRi are efficacious regardless of IGHV status

12%

major subsets

high propensity for Richter’s transformation

#2

#1

#4

#8

the paradigmatic indolent subset very aggressive

Major stereotyped subsets represent distinct CLL variants

Tobin et al. Blood 2003; Ghiotto et al. J Clin Invest 2004; Stamatopoulos et al. Blood 2007; Chu et al. Blood 2008; Catera et al. Mol Med 2008; Rossi et al. Clin Cancer Res 2009; Sutton et

• al. Blood 2009; Chu et al. Blood 2010; Marincevic et al. Haematologica 2010; Maura et al. PLosOne 2011; Ntoufa et al. Mol Med 2012; Agathangelidis et al. Blood 2012; Strefford et al.

Leukemia 2013; Rossi et al. Blood 2013; Papakonstantinou et al Mol Med 2013; Vardi et al. Clin Cancer Res 2013; Sutton et al. Mol Med 2014; Mansouri et al. J Exp Med 2015

very aggressive

Courtesy of K Stamatopoulos

BcR stereotypy refines prognostication in CLL

Distinct clinical outcomes for subsets #1, #2 and #4, independently of genomic aberrations or SHM status

Baliakas et al. Lancet Haematol. 2014

All express IGHV4-34, all concern IG mutated CLL, yet the outcome is different

Xochelli et al. Clin Cancer Res 2017

Subset #2 is as bad as CLL with TP53 aberrations though essentially devoid of such lesions

Baliakas et al. Blood 2015

Subset #8: the highest risk for Richter’s transformation amongst all CLL

Rossi et al. Clin Cancer Res. 2009 #4 #1 #2

5 10 15 20 25 30

Time years

0% 25% 50% 75% 100%

% untreated

#2, n=212 del(17p), n=284

p=0.5

Courtesy of K Stamatopoulos

TP53 abnormalities in CLL

5’ 3’ 1 DNA BINDING EX4 EX9 393 Missense Nonsense Frameshift

TP53

Frequency

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% MBL Early stage CLL CLL requiring treatment F-refactory CLL Richter syndrome TP53 M del17p13 TP53 M/del17p13 N=1/63 (1.5%) N=30/318 (9.4%) N=44/99 (44.4%) N=25/38 (65.7%) N=13/268 (4.8%) Dohner et al, New Engl J Med 2000 ; Rasi et al, Haematologica 2012; Zainuddin et al, Leuk Res 2011; Zenz et al J Clin Oncol 2010;Rossi et al Blood 2011; Rossi et al Blood 2014

Chr17

DNA damage P p53 P p21 cyclin B p53 P cyclin B cdc2 p21 BAX Caspase 9 Apoptosis Cell cycle arrest

M P

del(17p) p53wt (rare)

M P

del(17p) p53mut (>80%)

P P

LOH p53mut

M P

p53wt/mut (rare)

M P

normal p arm p53 q arm

17p-censored 11q-censored +12q-censored 13q-single- censored No aberration- censored

Months Hallek et al, Lancet 2010 0 6 12 18 24 30 36 42 48 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0 6 12 18 24 30 36 42 48 54 PFS Months

FCR

17p- on FCR

TP53 abnormalities in CLL

Stilgenbaueret al, Blood 2014 FC and TP53WT FC and TP53mut FCR and TP53WT FCR and TP53mut

Badoux Blood 2011; Fisher J Clin Oncol 2011; O’Brien, ASH 2014; Sharman ASH 2014; Byrd ASH 2015; Stilgenbauer, ASH 2015

Chemoimmunotherapy (CIT) vs novel agents in TP53 disrupted CLL

0% 20% 40% 60% 80% 100%

Response rate

0% 20% 40% 60% 80% 100%

CR PR PR-L

35% 7% 83% 78% 79%

12-months PFS

18% 22% 80% 79% 72%

Response rate PFS

CIT Novel agents CIT Novel agents

Relapsed/Refractory CLL

0.2 0.4 0.6 0.8 1.0 Proportion With PFS 6 12 18 24 30 36 42 Months del(17p) del(11q) No del(17p) or del(11q) + Censored

TP53 disruption is a prognostic biomarker in CLL treated with novel agents Ibrutinib in trials Venetoclax Idelalisib+R Ibrutinib in real-world practice

Byrd JC, Blood 2015; Thompson PA, Cancer 2015; Winqvist M, Haematologica 2016; Barrientos, ASCO,2015, 7011; Roberts, et al New Engl J Med 2016

Hallek et al. Blood. 2008; Oscier et al. Br J Haematol. 2013 Pospisilova et al. Leukemia. 2012 Zelenetz et al J Natl Cancer Inst 2015 Eichhorts et al, Ann Oncol 2015.

Guideline recommendations for TP53 analysis in clinical practice When What iwCLL Before treatment 17p deletion ERIC Before treatment TP53 mutation BCSH Before treatment 17p deletion and TP53 mutation NCCN Before treatment 17p deletion and TP53 mutation ESMO Before treatment 17p deletion and TP53 mutation

Thessaloniki Brno Uppsala Ulm London Novara Madrid Paris Amsterdam Copenhagen Bellinzona

Harmonization of TP53 mutation analysis

TP53 Guidelines TP53 Network TP53 Certification TP53 Manual

http://www.ericll.org/pages/networks/tp53network/ericmanualfortp53mutationalanalysis/!

a In patients who are not eligible for any other therapies

Chl: chlorambucil; CIRS: Cumulative Illness Rating Scale; Cr: creatinine

FCR Chl + anti-CD20

TP53

Chemo + anti-CD20

Age; CIRS; Cr clearance

BR Ibrutinib Idelalisib + Ra

IGHV

Ibrutinib Mutated and/or deleted Wild type Unmutated

Mutated

Biomarkers may inform first line treatment

Personal view

Outline

• The genetic landscape of CLL
• Molecular prognosticators of CLL
• Molecular predictors of CLL
• Richter syndrome biomarkers
• Novel biomarkers

BM PB Lymph node biopsy BIOPSY IS MANDATORY (PET-guided)

Clinical suspicion of RS

• Bulky disease
• Extranodal involvement
• B symptoms
• High LDH

Clinical clues of Richter transformation

NOTCH1 wt NOTCH1 M p<.001

• No. at Risk

NOTCH1 wt 531 279 92 31 11 3 1 NOTCH1 M 74 28 8 1 Events Total 5-year risk 95% CI NOTCH1 wt 18 531 3.9% 2.0-5.8 NOTCH1 M 12 74 18.6% 7.3-29.9

• No. at Risk

NOTCH1 wt & no IGHV4-39 519 273 90 30 11 3 1 NOTCH1 wt & IGHV4-39 12 12 12 12 NOTCH1 M & no IGHV4-39 67 27 8 1 NOTCH1 M & IGHV4-39 7 1 Events Total 5-year risk 95% CI NOTCH1 wt & no IGHV4-39 18 519 4.0% 2.1-5.9 NOTCH1 wt & IGHV4-39 12 NOTCH1 M & no IGHV4-39 8 67 12.5% 2.9-22.1 NOTCH1 M & IGHV4-39 4 7 75.0% 32.5-100

NOTCH1 wt & no IGHV4-39 NOTCH1 M & no IGHV4-39 NOTCH1 M & IGHV4-39 NOTCH1 wt & IGHV4-39 p<.001 p<.001

Risk of Richter transformation according to NOTCH1 mutation status and IGHV usage at CLL diagnosis

MYC activation TP53 disruption

CLL DLBCL (Richter)

transformation

MYC translocation amplification NOTCH1 mutations

TP53 disruption, MYC activation and CDKN2A loss contribute to CLL transformation to Richter syndrome

MGA mutations

Fabbri, et al. J Exp Med 2011 Rossi, et al. Blood 2011 Rossi, et al. Blood 2012 Rossi, et al. Leuk Lymphoma 2012 Chigrinova et al, Blood, in press

CDKN2A loss

Clonal relationship of Richter syndrome

CLL

80% 20%

Clonally related Richter Clonally unrelated Richter

V4-39 D6 J4 V4-39 D6 J4 V4-34 D2-2 J3

The genetic profile of clonally unrelated RS differs from that of clonally related RS

p=.018 p=.009

Clonally unrelated Clonally related

p=.017 Rossi et al, Blood 2011; Rossi and Gaidano, 2018

Outline

• The genetic landscape of CLL
• Molecular prognosticators of CLL
• Molecular predictors of CLL
• Richter syndrome biomarkers
• Novel biomarkers

Thompson PA, ASH 2014, Oral Presentation #22

Complex Karyotype in the novel agent era

Ibrutinib MDACC Ibrutinib PCYC-1102/1103

O’Brien S, et al. ASH 2016; 233

Venetoclax

Anderson MA, Blood 2017

2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 2 2 2 4 2 6 2 8 3 0 3 2 3 4 3 6 3 8 4 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 1 0 0

T im e (M o n th s ) P ro b a b ility o f P F S (% )

C K T & E ith e r d e l(1 7 p )/T P 5 3 -M (n = 1 6) C K T & N e ith e r d e l(1 7 p )/T P 5 3 -M (n = 1 0) N o C K T & E ith e r d e l(1 7 p )/T P 5 3 -M (n = 1 7) N o C K T & N e ith e r d e l(1 7 p )/T P 5 3 -M (n = 2 2)

Idelalisib-R GS 0116/0117

Kreuzer, et al. Abstract #192, ASH, 2016.

BCR

CD79A

CARD11

CD79B

BTK

MALT1 BCL10 IKBKB

PLCG2

Non-canonical NF-κB pathway

TRAF3

TRAF2 MAP3K14

IKK

BIRC3

NF-kB activation

Ibrutinib

Molecular mechanisms of resistance to ibrutinib

BCR pathway

CLL DLBCL MCL In CLL: BTK C481S mutation PLCG2 mutation Absent in ibrutinib naïve patients

Davis, Nature 2010 Woyach, NEJM 2014 Furman, NEJM 2014 Famà, Blood 2014 Rahal, Nat Med 2014

Clonal architecture of TP53 mutated CLL Scenario 2 Scenario 3 Scenario 1

Nadeu F et al Blood 2016

subclonal M clonal M+subclonal M clonal M wt

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Allele frequency corrected for tumor representation

TP53 mutations Sanger sequencing negative n=50 Sanger sequencing positive n=35

TP53 mutations

Small TP53 mutated subclones in untreated CLL

85%

(N=263)

N=309

6%

(N=18)

3%

(N=10)

6%

(N=18)

Ultra deep-NGS 85 TP53 mutations in 46/309 (15%) CLL

* TP53 mutations and 17p13 deletion Median allele frequency: 2.1% (range: 0.3-11%)

TP53 unmutated Solely subclonal TP53 M Clonal TP53 M

• .0042

.<.0001 .0042

• .6926

<.0001 .6926

• p from pairwise comparisons

Rossi, Blood 2014

INCREASING CCF TP53 subclone evolution in CLL treated within the CLL8 trial

Landau, Nature 2015

Selection of the TP53 mutated subclone Expansion of the TP53 mutated clone

Diagnosis Chemotherapy Progression Chemotherapy Refractoriness

TP53 mutated CLL cell

NGS Sanger

Rossi, Blood 2014; Malcikova, Leukemia 2015; Landau, Nature 2015; Nadeu, Blood 2016

Summing up

• CLL is characterized by a marked degree of molecular heterogeneity,

since few mutations recur patients at a frequency >5%

• TP53 disruption identifies a genetic category of high risk CLL,

predicts chemoimmunotherapy failure and mandates treatment with innovative drugs, including ibrutinib, idelalisib or venetoclax

• Mutated IGHV gene represent a predictive biomarker for identifying

patients that may benefit the most from chemoimmunotherapy with FCR

• Novel molecular prognosticators and predictors are under scrutiny,

but their application in the clinical practice is premature

Lodovico Terzi di Bergamo Valeria Spina Davide Rossi Fary Diop Riccardo Moia Lorenzo De Paoli Gloria Margiotta Casaluci Clara Deambrogi Chiara Favini Ahad Ahmed Kodipad Sruthi Sagiraju Nawwar Maher Simone Favini Ilaria Del Giudice Sabina Chiaretti Monica Messina Anna Guarini Robin Foà

CRO AVIANO

Michele Da Bo Valter Gattei Roberto Marasca Carol Moreno Giovanni Del Poeta Silvia Deaglio Francesco Forconi

Grant support:

Luca Laurenti

CRO AVIANO

Paolo Ghia Laura Pasqualucci Giulia Fabbri Riccardo Dalla Favera

DSL NOTCH1 ICN S3 γ-secretase S2 Metalloprotease Pro-NOTCH ICN ER-Golgi RPBJ MAML Other Co-Activators gene expression Degradation SPEN RPBJ DTX1 Ubiquitination Degradation

5’ 3’

EGF repeats (1-36) LNR RAM HD TM Ankyrin

1 2556 2155 2556

CLL Missense Nonsense Frameshift

TAD PEST

TAD PEST

NOTCH1 mutations in CLL

0% 10% 20% 30% 40% 50%

Frequency (%)

N=60/539 (11%) N=18/58 (31%) N=10/48 (20%) N=2/134 (1%)

*** *** P<0.001 *** P<0.05 ** **

N=2/63 (3%)

de novo DLBCL MBL CLL diagnosis Richter syndrome

**

F-ref CLL

Arruga F et al. Leukemia 2013 Arruga F et al. Leukemia 2016 Fabbri G et al. PNAS 2017 Pozzo F et al. Leukemia 2017 Fabbri, et al. J Exp Med 2011 Puente, et al. Nature 2011 Wang, et al. New Engl J Med 2011 Rossi, et al. Blood 2012 Rasi, et al. Haematologica 2012

MYC (proliferation) DUSP22 (migration) CD20 (anti CD20)

3’ UTR

NOTCH1 mutations as predictive marker for no benefit from addition of rituximab to chemotherapy

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90 96

PFS Months

Stilgenbauer S et al. Blood 2013

GCLLSG CLL8

PFS Months

GCLLSG CLL11

Estenfelder S et al. Blood 2016 128:3227