Complex karyotype in CLL Arnon Kater, Panagiotis Baliakas, Alexander Leeksma MD On behalf of the ERIC ”complex karyotype project” Good afternoon everyone, today I will give a short update of the complex karyotype project in CLL on behalf of the ERIC complex karyotype working group.

Introduction Accepted diagnostic workup CLL prior to treatment FISH 13q, tris 12, 11q and 17p; TP53 mutation according to ERIC guidelines Since 2006 data individual groups of predicitive value of complex karyotype (CK), never integrated in guidelines New interest in CK in era of novel agents (Thompson, Cancer 2015)

Questions How to define CK? Predictive value of CK? Differences in methodology? Prospective analysis? ERIC wide registry Group of dedicated researches First meeting at MLL Munich fall 2016 Grant Janssen

Overview Short summary results complex karyotype (CK) classical cytogenetics study (on behalf of Panagiotis Baliakas/Kostas Stamatopoulos) Preliminary results CK CGH array study Because Panagiotis is not attending this meeting I will give a short summary of the data of the ERIC classical cytogenetics study. After the first part I will show you some preliminary results of the CGH array study

Overview of the cohort 3656 3580 ≥10 metaphases N=3580 N, % Male Female 2252, 63% 1328, 37% Median age diagnosis <55 >70 65.6 years 698/3560, 20% 1184/3560, 33% MBL Binet A Binet B Binet C 258/2863, 9% 2098/2893, 73% 357/2893, 12% 258/2893, 9% M-CLL 1222/2051, 60% TP53abs 299/3308, 9% del(11q) 377/3256, 12% Trisomy 12 507/3260, 16% del(13q) 1769/3271, 54% Treated (median FU: 4.6 years) 1413/3393, 42% 3656 3580 ≥10 metaphases This is an overview of the cohort consisting of more than 3000 CLL patients. As you can see it is a reflexion of a normal cohort.

Method and timing of testing CPG/IL2 n=379, 11% TPA, n=1846, 52% TPA, CPG+IL2 n=1355, 37% 3580 Tested within 6 months from diagnosis N=2684/3549, 76% Tested within 1 year from diagnosis N=2814/3549, 79% Tested before treatment (treated cases) N=1238/1406, n=88%

Overview of chromosomal aberrations 3580 3179 (89%) Non-CK 741 (23%) 1 aberration [non-del(13q)] 406 (13%) 2 aberrations 401 (11%) CK 2032 (64%) Normal/del(13q) No difference between cell stimulation protocols

CK (≥3 aberrations) vs non CK P-value Male 275/401, 69% 1977/3179, 63% 0.012 <55 years 81/397, 20% 617/3163, 20% 0.67 >70 years 144/397, 36% 1040/3163, 33% 0.17 MBL Binet A Binet B Binet C 13/295, 4% 186/295, 63% 66/295, 22% 30/295, 10% 245/2568, 10% 1912/2568, 74% 291/2568, 11% 120/2568, 5% 0.003 <0.0001 M-CLL 98/254, 38% 1124/1797, 63% TP53abs 117/381, 31% 182/2927, 6% del(11q) 106/3691, 29% 271/2887, 9% Trisomy 12 98/3682, 27% 409/2892, 14% del(13q) 237/3913, 60% 1532/2900, 53% 1: 14/105, 13% TP53abs, 2: 17/97 TP53abs, 39/97: +12,+19, 3: 71/235 (30%) negative for TP53abs, del(11q), +12

CK and OS Selected cases karyotyped before treatment We first looked at overall survival in complex vs non-CK patients. The definition for a complex CLL is 3 or more chrom aberrations 13q excluded. As expected complex karyotype is associated with worse survival Selected cases karyotyped before treatment

CK and TP53mut Multivariate analysis CK Age Mutation TP53 If we go further into this and look at te effect of a p53 mutation you can see that this is still the case with a p53 mut. Multivariate analysis CK Age Mutation TP53

3abs vs 4abs vs ≥5abs Complex karyotypes with more abs are doing worse than the ones with less abs.

Are all CK equal? Interestingly a complex karyotype with trisomy 12 and 19 can be seen as a prognostically favorable subgroup. The question is of course if we should make another risk stratification based on this result.

Preliminary conclusions Classical cytogenetics (retrospective analysis) Different mitogens homogeneous results CK independent risk factor but: different subgroups can be found (e.g. +12,+19) exact definition of CK to be determined Submitted to IWCLL on behalf of ERIC

Overview short summary results complex karyotype (CK) classical cytogenetics study (on behalf of Panagiotis Baliakas/Kostas Stamatopoulos) preliminary results CK CGH array study Because Panagiotis is not attending this meeting I will give a short summary of the data of the ERIC classical cytogenetics study. After the first part I will show you some preliminary results of the CGH array study

Preliminary results CK CGH array study (first analysis) ERIC survey: 25 centers will provide data (different sizes) Very preliminary results of first 9 centers Methods: For del13q, del11q, del17p: no cut-off All other abberations: cut-off of >5MB (Schoumans diagnostic guidelines) 1. effect complex karyotype (CK) on survival 2. effect different aberrations on survival As promised I will show you some preliminary results of the CK CGH array study. Thus far we only have the results of 9 out of the 25 centers involved in this large study. The first analysis we performed is about the CGH array data only.

Overview of the CGH array cohort N=1322 N, % Male Female 892, 67,5% 427, 32,3% Median age diagnosis <55 >70 63 years 146/677, 22% 168/677, 25% MBL Binet A Binet B Binet C 13/594, 2% 432/594, 73% 102/594, 17% 47/594, 8% M-CLL 311/554, 56% TP53abs 55/618, 9% del(11q) 182/1294, 14% Trisomy 12 174/1294, 13% del(13q) 584/1293, 45 % 1322 This is an overview of the CGH array cohort. The cohort is lower in number but still a good reflexion of the general population of CLL patients.

Overview of chromosomal aberrations 1322 1148 (87%) Non-CK 552 (42%) 1 aberration 191 (14%) 2 aberrations 174 (13%) CK 405 (31%) Normal This is a schematic overview of the subdivision of the CK and non-CK based on CGH arrays. The CK percentages are a little bit higher compared to the CK status defined by classical cytogenetics in the other cohort.

Overview effect recurrent aberrations The cohort is in line with earlier studies which show an effect of del11q and 17p on overall survival. For del13q and trisomy 12we could not find any differences in overall survival although there is a trend towards a lower overall survival if you look at trisomy 12.

Overview effect other aberrations Besides the recurrent aberrations we looked with CGH arrays at other less recurrent aberrations not regularly tested/included/analyzed by the standard FISH panels. Trisomy 18 and 19 alone do not have an effect on survival as well as duplication of 8q. A duplication of 2p and especially a deletion of 6q seem associated with an inferior prognosis.

Dissecting CK If we dissect the CK in this cohort we get this pie chart.

Dissecting CK For the array analysis we see the same picture if we compare non-CK with complex karyotypes if we look at overall survival.

Dissecting CK If we subdivide te complex cases we see that especially 5 or more aberrations are associated with a worse survival while 3 or 4abs are somewhere in between.

Effect TP53mutation And if we compare non complex karyotype with complex karyotype in the presence of a p53 mutation you can see that the CK patients do worse.

Preliminary conclusions Array-based (work in progress) Complex karyotype is a prognostic factor in CLL The presence of ≥5abs rather than >3abs is associated with the worst clinical outcome Genomic complexity is associated with worse clinical outcome even amongst cases harboring TP53mut

Acknowledgements ERIC CK working group Sabine Franke, Liege, Belgium Panagiotis Baliakas, Uppsala , Sweden Richard Rosenquist, Uppsala, Sweden Sabine Franke, Liege, Belgium Anh Nhi Tran, Stockholm, Sweden Florence Nguyen-khac, Paris, France Claudia Haferlach, München, Germany Loïc Ysebaert, Toulouse, France Karl-Anton Kreuzer, Cologne Germany David Oscier, Bournemouth, Great Britain Douka Vasiliki, Thessaloniki, Greece Constantine Tam, Melbourne, Australia Kostas Stamatopoulous, Thessaloniki, Greece Maria Laura Blanco, Barcelona, Spain Hidde Posthuma, Amsterdam, the Netherlands Anna Puiggros, Barcelona, Spain Clemens Mellink, Amsterdam, the Netherlands Blanca Espinet, Barcelona, Spain Arnon Kater, Amsterdam, the Netherlands Jacqueline Schoumans, Lausanne, Switzerland Ana Eugenia Rodriguez, Salamanca, Spain Sarka Pospisilova, Brno, Czech Republic Karla Pevlova, Brno, Czech Republic Kaja Malinka, Brno, Czech Republic Sabine Jeromin, München, Germany Marian Stevens-Kroef, Nijmegen, the Netherlands I would like to thank all the members of the complex karyotype working group. It is really a pleasure to work with you all. Thank you very much for your attention. I am open to answer questions. (If you are not on board and want to be part of this large initiative to address the impact of genomic complexity in CLL please send an email to a.c.leeksma@amc.uva.nl)

CK and IGHV status (classical cytogenetics

CK and IGHV status (classical cytogenetics)

CK and IGHV status (CGH array)

Dissecting CK (CGH array) If we compare IGHV UM and look at the effect of a complex karyotype we do not see any difference if we compare UM cases non complex with a complex karyotype. This means that mutation status in CLL is a better prognostic marker than an complex karyotype.

CK and IGHV status (classical cytogenetics)

Are all TP53mut equal? noTP53abs, n=2974

Are all TP53mut equal? noTP53abs, n=2974

Dissecting CK This picture does not change if we change the criteria for a complex karyotype when we do not include 13q deletions as aberrant.

Effect of IGHV The complex karyotypes are seen much more frequently in UM CLL and in UM CLL the CKs are associated with a worse survival compared to the M-CLL patients.

Effect TP53 mutation As expected for p53 mutations we see that in the absence of a functional p53 survival is worse compared to a normal p53.

Effect TP53 mutation But if we look at complex karyotypes and look at the effect of a p53 mutation we see that within the complex karyotypes p53 mutations still do worse.

CGH array vs FISH recurrent aberrations Male Female 177, 69% 79, 31% Median age diagnosis <55 >70 65 years 43/248, 17% 86/248, 35% MBL Binet A Binet B Binet C 13/209, 6% 146/209, 70% 32/209, 15% 18/209, 9% M-CLL 47/119, 39% aberration CGH array FISH 13qmono 110 122 13qbi 29 21 trisomy12 37 38 del11q 52 56 del17p 16 22 Total 244 259 This a sneak preview of the next part of the CGH array project. The idea is to compare CGH array with classical cytogenetics and FISH. Some labs nowadays only use CGH instead of CK in combination with FISH. The question is or FISH is still useful or that we can do without. In the first analyzed patients we see some differences. With FISH we detect a little bit more aberrations if we compare this with CGH array without a size cut-off. We need more data to say something about these differences.