(2014) 28, 1511–1518 & 2014 Macmillan Publishers Limited All rights reserved 0887-6924/14 www.nature.com/leu

ORIGINAL ARTICLE High hyperdiploidy among adolescents and adults with acute lymphoblastic leukaemia (ALL): cytogenetic features, clinical characteristics and outcome

L Chilton1, G Buck2, CJ Harrison1, RP Ketterling3, JM Rowe4, MS Tallman5, AH Goldstone6, AK Fielding7 and AV Moorman1

High hyperdiploidy (HeH, 51–65 chromosomes) is an established genetic subtype of acute lymphoblastic leukaemia (ALL). The clinical and cytogenetic features as well as outcome of HeH among adolescents and adults have not been thoroughly investigated. Among 1232 B-cell precursor ALL patients (15–65 years) treated in the UKALLXII/ECOG2993 trial, 160 (13%) had a HeH , including 80 patients aged 424 years. The frequency of HeH was the same in Philadelphia chromosome (Ph)-positive and -negative cases, but Ph-positive patients were older. The cytogenetic profiles of Ph-positive and Ph-negative HeH cases were similar, although 2 was strongly associated with Ph-positive HeH. Overall, Ph-positive HeH patients did not have an inferior overall survival compared with Ph-negative patients (P ¼ 0.2: 50 vs 57% at 5 years). Trisomy of was associated with a superior outcome in Ph-negative patients, whereas þ 5 and þ 20 were associated with an inferior outcome in Ph-positive and Ph-negative patients, respectively. All three markers retained significance in multivariate analysis adjusting for age and white cell count: hazard ratio for risk of death 0.47 (95% CI: 0.27–0.84) (P ¼ 0.01), 3.73 (1.51–9.21) (P ¼ 0.004) and 2.63 (1.25–5.54) (P ¼ 0.01), respectively. In conclusion, HeH is an important subtype of ALL at all ages and displays outcome heterogeneity according to chromosomal gain.

Leukemia (2014) 28, 1511–1518; doi:10.1038/leu.2013.379 Keywords: acute lymphoblastic leukaemia; high hyperdiploidy; adults; adolescents; prognosis; chromosomal abnormalities

INTRODUCTION frequency of Ph-positive ALL is 10–15 times higher than that in High hyperdiploidy (51–65 chromosomes, HeH) is an established childhood ALL, such cases are more prevalent. In contrast to HeH, genetic subtype of B-cell precursor acute lymphoblastic leukaemia patients with Ph-positive ALL have a very poor outcome when (BCP-ALL), which is associated with a favourable prognosis, treated with conventional . However, the advent of especially in children.1 HeH is observed in 40% of young inhibitors has improved the outcome of these 1 children (1–4 years) with BCP-ALL, and decreases in frequency patients markedly. The coexistence of two established genetic with age to 10% among adults (25–59 years).1 The underlying abnormalities with opposing prognostic associations raises molecular consequence of the massive chromosomal gain that important clinical and biological questions. We recently reported defines HeH is unclear, but clonal stability between diagnosis and that HeH patients with concurrent chromosomal translocations, 10 relapse and the back-tracking of the HeH clone to birth indicate including t(9;22), had distinctive patterns of chromosomal gain. that it is an early and driving genetic event.2,3 The pattern of The UKALLXII/ECOG2993 trial recruited the largest single cohort of chromosomal gain is non-random, with eight chromosomes (X, 4, adolescents and adults (15–65 years) with ALL who were treated 11 6, 10, 14, 17, 18 and 21) accounting for the vast majority of gains.4 uniformly. Our previous studies on this cohort have demonstrated Numerous paediatric studies have demonstrated outcome that among Ph-negative patients those with HeH had significantly heterogeneity on the basis of various cytogenetic characteristics— superior survival and that the same trend was also observed among 12,13 for example, modal chromosome number (MCN) and the presence adults with Ph-positive disease. In this report, we focus on the of specific .5 Although the results have not always been 160 adolescents and adults with HeH ALL, including 48 with consistent, higher MCN and trisomies of chromosomes 4, 10, 17 and Ph-positive HeH ALL treated in the UKALLXII/ECOG2993 trial. 18 have been associated with a decreased risk of relapse within this We describe their cytogenetic profile, compare Ph-positive and subgroup.6–9 The prognostic effect of specific trisomies and MCN Ph-negative HeH patients and, for the first time in adult ALL, has not been studied previously in adult ALL. examine the prognostic impact of MCN and specific trisomies. The coexistence of two primary genetic aberrations within the same clone is rare in ALL, but the concurrent presence of HeH and t(9;22)(q34;q11)/Ph is an exception. The frequency of this PATIENTS AND MATERIALS ‘double-hit’ is very low in paediatric ALL because the incidence of All patients included in this study were diagnosed with ALL and Ph-positive ALL is only 2%. However, in adult ALL, in which the treated in the UKALLXII/ECOG2993 trial between 1993 and 2008 as

1Leukaemia Research Group, Northern Institute for Research, Newcastle University, Newcastle upon Tyne, UK; 2Clinical Trial Service Unit, University of Oxford, Oxford, UK; 3Department of Cytogenetics, Mayo Clinic, Rochester, MN, USA; 4Department of Hematology, Shaare Zedek Medical Center, Jerusalem, Israel; 5Memorial Sloan Kettering Cancer Center, Weill Cornell Medical College, New York, NY, USA; 6Department of Haematology, University College Hospital London, London, UK and 7Department of Haematology, Royal Free and University College of London Medical School, London, UK. Correspondence: Professor AV Moorman, Leukaemia Research Cytogenetics Group, Northern Institute for Cancer Research, Newcastle University, Level 5, Sir James Spence Institute, Royal Victoria Infirmary, Newcastle upon Tyne NE1 4LP, UK. E-mail: [email protected] Received 21 October 2013; revised 3 December 2013; accepted 5 December 2013; accepted article preview online 19 December 2013; advance online publication, 17 January 2014 High hyperdiploidy in adult ALL L Chilton et al 1512 previously described.13,14 The Ethics Committee or Institutional Review Board of each participating centre approved the study. Informed Table 1. Demographic and clinical features, treatment and outcome consent was obtained from all subjects in accordance with the for 160 ALL patients with a HeH karyotype (51–65 chromosomes), Declaration of Helsinki. Full treatment details and overall results of the stratified by the presence or absence of the Ph trial have been published.11,14 Patients received two phases of standard induction therapy. Ph-negative patients with an HLA-matched sibling Total, n (%) Ph-positive, Ph-negative, P donor were assigned to receive an allogeneic transplant, whereas those n (%) n (%) without a donor were randomly assigned to receive either standard Total number of patients 160 (100) 48 (100) 112 (100) consolidation/maintenance chemotherapy or a single autologous transplantation. Ph-positive patients received either a sibling-matched Sex allogeneic transplant or a matched unrelated donor transplant. The Male 97 (61) 28 (58) 69 (62) 0.7 tyrosine kinase inhibitor, , was made available to Ph-positive Female 63 (39) 20 (42) 43 (38) patients treated in the UKALLXII/ECOG2993 trial in 2003, and 22/48 Age (46%) of the Ph-positive patients in this study received imatinib during 15–24 80 (50) 13 (27) 67 (60) 0.001 the course of their therapy. Full details of the imatinib therapy delivered 25–39 31 (19) 12 (25) 19 (17) to these patients have been published.14 Briefly, eligible patients 40–65 49 (31) 23 (48) 26 (23) received 400–600 mg of imatinib per day for 1 month either post 9 a induction or during the second phase of induction. Ph-positive patients WCC ( Â 10 /l) o30 131 (82) 32 (67) 99 (88) 0.002 who achieved a complete remission (CR) after induction were eligible for 30–99.9 19 (12) 12 (25) 7 (6) transplant as described above. X100 9 (6) 4 (8) 5 (4) Cytogenetic analysis of pre-treatment bone marrow or peripheral blood was performed locally, reviewed centrally and collated Bone marrow blasts (%)b retrospectively as previously described.13 Patients with a T-ALL o90 38 (24) 7 (15) 31 (28) 0.8 immunophenotype were not included in this present analysis. All 90–94 38 (24) 7 (15) 31 (28) 95–99 38 (24) 9 (19) 29 (26) patients were classified as having HeH (51–65 chromosomes) either on X100 9 (6) 1 (2) 8 (7) the basis of cytogenetic analysis (n ¼ 149) or by fluorescence in situ hybridisation (n ¼ 11). Patients classified as HeH on the basis of CR fluorescence in situ hybridisation must have showed evidence for the Yes 152 (95) 45 (94) 107 (96) 0.7 gain of 5–19 chromosomes (equivalent to a modal number of 51–65 No 8 (5) 3 (6) 5 (4) chromosomes) with no evidence of triploidy or tetraploidy. Fluorescence Relapse in situ hybridisation was performed on patients with a normal karyotype Yes 46 (30) 14 (31) 32 (30) — or with failed cytogenetics (where material was available) and was No 106 (70) 31 (69) 75 (70) undertaken using either the Chromoprobe Multiprobe-I system (Cytocell, Banbury, Oxon, UK), which enumerated all 24 chromosomes, or with First remission death multiple centromeric probes as previously described.15 Yes 30 (20) 13 (29) 17 (15) — Event-free survival (EFS) was defined as the time from the start of No 122 (80) 32 (71) 90 (80) treatment to relapse or death and overall survival (OS) was defined as the Death time from the start of treatment to death. Patients without an event of Yes 77 (48) 26 (54) 51 (46) — interest were censored at the date of last contact. Relapse rate (RR) was No 83 (52) 22 (46) 61 (54) calculated only for patients who achieved a CR, and was defined as the time from the start of treatment until relapse, with deaths in remission First remission treatment Chemo/Auto 98 (64) 19 (42) 79 (74) o0.001 being censored. CR was defined as a bone marrow aspirate with o5% Allo/MUD 54 (36) 26 (58) 28 (26) leukaemic blasts and evidence of regeneration of normal hematopoietic cells. The median follow-up time for EFS and OS was 7.7 years and 7.6 Rate at 5 years (95% CIs) years, respectively. EFS, RR and OS estimates were calculated using the EFSa 48% (40–56) 42% (27–55) 51% (41–60) 0.07 Kaplan–Meier method. Among Ph-negative patients survival between RR 34% (27–44) 33% (20–51) 35% (26–45) 0.5 genetic subgroups was calculated using univariate Cox models. For Ph- OS 55% (47–62) 50% (35–63) 57% (47–66) 0.2 positive patients, a variable for treatment with imatinib was included. Abbreviations: ALL, acute lymphoblastic leukaemia; Allo, sibling donor Multivariate Cox regression modelling was performed for RFS, EFS and OS allogenic transplant; Auto, autologous transplant; Chemo, chemotherapy; using a forward-selection stepwise modelling process with a base model of CI, confidence interval; CR, complete remission; HeH, high hyperdiploidy; age (as a continuous variable) and log-transformed initial white blood cell MUD, matched unrelated donor; OS, overall survival; Ph, Philadelphia count (WCC) factor and the likelihood-ratio test to compare models. The chromosome; RR, relapse rate; WCC, white cell count. aData not available proportional hazard assumption of the final models was assessed by visual for one patient. bData unavailable for 37 patients. inspection of the log-log plot and Grambsch and Therneau’s proportional hazards test. Other comparisons were performed using the w2 test, Fisher’s exact test or Wilcoxon rank-sum test, as appropriate. Due to the Cytogenetic features of HeH cases with and without the Ph investigative nature of this analysis, we did not apply a stringent multiple comparisons adjustment. However, all tests were conducted at the 1% The cytogenetic features of HeH patients were investigated in a significance level. All analyses were performed using Intercooled Stata 12.1 cohort of 137 cases with complete . The MCN for the for Windows (Stata Corporation, College Station, TX, USA). entire cohort was 55 and this was the same for Ph-positive and Ph-negative cases (Figure 1a). However, the MCN distribution was slightly skewed towards lower MCN for Ph-positive patients and they were marginally more likely to have an MCN of 51–54 RESULTS compared with Ph-negative patients: 50 vs 33% (P ¼ 0.05). Overall, Frequency and demographic features of HeH in adult ALL 45 (33%) cases had solely numerical abnormalities, whereas 92 Among 1232 appropriately tested BCP-ALL patients treated in the (67%) had one or more structural aberrations, including 48 Ph- UKALLXII/ECOG2993 trial, 160 (13%) had a HeH karyotype. The positive patients. The t(9;22) was the sole structural abnormality in frequency of HeH was similar among Ph-positive (48/340, 14%) 34 (71%) Ph-positive patients, whereas 14 (29%) harboured additional and Ph-negative (112/892, 13%) patients. Overall, the median age structural abnormalities, predominantly abnormalities of chromosome of HeH patients was 25 years but Ph-positive patients were arm 1q (n ¼ 5) or chromosome 19 (n ¼ 4). Among the 44 Ph-negative significantly older than Ph-negative patients: median age 37 vs 20 patients with structural aberrations, abnormalities of years, respectively (Table 1). Ph-positive HeH patients also had a were the most prevalent (n ¼ 11), which included deletions (n ¼ 8) higher WCC compared with Ph-negative HeH patients: median and partial duplications (n ¼ 3) of 1q. Other recurrent abnormalities 19.5 vs 4 Â 109/l, respectively (Table 1). included isochromosome of 17q (i(17)(q10)) (n ¼ 4), of

Leukemia (2014) 1511 – 1518 & 2014 Macmillan Publishers Limited High hyperdiploidy in adult ALL L Chilton et al 1513 20 compared with 38 and 22% for Ph-positive patients treated with Ph-negative (n=89) 11,12,14 18 and without imatinib. The vast majority of HeH patients Ph-positive (n=48) achieved a CR (152/160, 95%) but 46 (30%) patients went on to 16 relapse and a further 30 (20%) died in first remission (Table 1). 14 The OS rate at 5 years was 55%, which is greater than that 12 reported for the entire cohort. We have previously reported that, 10 among Ph-negative patients on this trial, those with HeH had a significantly improved EFS and OS.13 In addition, Ph-positive HeH 8 patients had a significantly reduced risk of relapse when 6 Number of cases compared with other Ph-positive patients, but only a borderline 12 4 improved OS and EFS. Table 1 compares the outcome of HeH 2 patients by Ph status and shows that Ph-positive patients did not have an inferior CR rate (Fisher exact test, P ¼ 0.7) or an increased 0 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 RR at 5 years compared with Ph-negative patients (log rank Modal Number P ¼ 0.5, Figure 2a). The RR at 5 years was not significantly different between Ph-positive and Ph-negative patients even if the patients 100% Ph-negative (n= 89) who received an allogenic transplant in first remission were Ph-positive (n=48) excluded (48% (95% CI: 26–75%) vs 43% (95% CI: 33–56%), 90% P ¼ 0.1). In addition the median time to relapse was similar in the 80% * two subgroups (Ph-negative 15.5 vs Ph-positive 11.4 months, 70% Mann–Whitney test P ¼ 0.5). Ph-positive patients appeared to * have a higher rate of death in first remission compared with 60% Ph-negative patients, but this did not reach statistical significance 50% at the 1% level (log rank analysis at 5 years 34% (95% CI: 21–51%) vs 18% (95% CI: 12–28%), P ¼ 0.04). This marginal difference is not 40% surprising given the differences in transplant eligibility between 30% the two subgroups of patients (see methods). Transplant-related Cases with gains (%) 20% mortality was the main cause of death among Ph-positive (8/13) and Ph-negative (7/17) patients. Among the non-transplanted 10% patients, infection was the principal specified cause of death 0% during first remission (seven patients). The overall outcome of 1 2 3 4 5 6 7 8 9 X Y

10 11 12 13 14 15 16 17 18 19 20 21 22 Ph-positive patients was not inferior compared with that of Ph-negative patients: hazard ratio for risk of death 1.38 (95% CI: der(22) Chromosomes Gained 0.86–2.21), P ¼ 0.2 (Table 1, Figure 2b). However, 22 Ph-positive patients received with imatinib, which we Figure 1. Distribution of MCN (a) and pattern of chromosomal gain have recently reported as increasing long-term survival.14 (b) for 137 adolescents and adults with high hyperdiploid (HeH, 51– The equivalent hazard ratio, excluding Ph-positive patients 65 chromosomes) ALL stratified by Ph status. Cases in which the treated with imatinib, showed a marginally significant inferior HeH karyotype was identified by fluorescence in situ hybridisation OS for Ph-positive patients: 1.85(1.08–3.17), P 0.03. However, a (n ¼ 11) or with an incomplete karyotype (n ¼ 12) were not included ¼ in these figures. Symbols: *Po0.01. multivariate Cox regression model adjusting for age, WCC and first remission treatment showed no difference in the hazard ratio of OS between Ph-negative and Ph-positive patients whether or not 6q (n ¼ 3) and translocations involving the IGH@ locus: t(14;18)(q32;q21)/IGH@-BCL2, t(14;19)(q32;q13)/IGH@-CEPBA and patients treated with imatinib were included in the model: 0.72 t(8;14)(q11;q32)/IGH@-CEPBD (one case each). (95% CI: 0.40–1.28), P ¼ 0.3 and 0.96 (95% CI: 0.52–1.78), P ¼ 0.9, The pattern of chromosomal gains was clearly non-random, respectively. with chromosomes 4, 6, 10, 14, 18, 21 and X gained in more than 50% of cases (Figure 1b). In most instances, one extra copy of the Outcome heterogeneity by cytogenetic features chromosome was gained. However, two extra copies () of were observed in 58/137 (42%) cases, being In order to investigate the relationship between individual more prevalent among Ph-positive cases: 46/89 (52%) vs chromosomes and outcome, we compared the outcome of patients with and without a given chromosomal gain using a 12/48(25%), P ¼ 0.003. There were no significant differences in the pattern of trisomies and tetrasomies by sex or age. However, univariate Cox regression model and assessed three end points— there were differences in the pattern of gain between Ph-positive EFS, RR and OS. In the Ph-negative cohort, almost all chromo- X and Ph-negative cases (Figure 1b). The most striking difference somes were gained in a sufficient number of cases (n 10) to be was the frequency of trisomy 2, which was observed in 65% of assessed and eight chromosomes showed some evidence (Po0.1) of an association with one or more of the end points (Table 2). Six Ph-positive cases but only in 12% of Ph-negative cases (Po0.001). Although trisomy of chromosomes 1, 10, 12, 17 and 18 appeared chromosomes showed some evidence of an association with a to be more common among Ph-negative patients, the difference superior outcome ( þ 4, þ 10, þ 14, þ 17, þ 21 and þ X), whereas did not reach statistical significance. An extra copy of the Ph two chromosomes were linked to an inferior outcome ( þ 5, þ 20). chromosome (der(22)t(9;22)(q34;q11)) was present in 38/48 (79%) In the light of studies by the Children’s Group we Ph-positive patients, which was much more frequent than the examined the prognostic impact of double trisomy ( þ 4, þ 10) gain of a normal in Ph-negative patients: 33/109 (n ¼ 50) and triple trisomy ( þ 4, þ 10, þ 17) (n ¼ 32). Both double trisomy and triple trisomy patients were associated with a (30%) (Po0.001). significantly superior OS compared with Ph-negative HeH patients without those combinations: hazard ratio for OS ¼ 0.42 (95% CI: Outcome of HeH patients with and without a Ph 0.23–0.77) (P ¼ 0.005) and 0.47 (0.23–0.95) (P ¼ 0.034), respectively. The OS of all patients treated in the UKALLXII/ECOG2993 trial However, when we compared þ 4 patients with and without was 39% at 5 years.11 Ph-negative patients had an OS of 43% double trisomy or triple trisomy no difference in OS was seen

& 2014 Macmillan Publishers Limited Leukemia (2014) 1511 – 1518 1514 ( So hpstv n ngtv ains ( patients. -negative and Ph-positive of OS ekma(04 51–1518 – 1511 (2014) Leukemia 2. Figure c ainswtotteseie anwt epc oters fa vn,rlpeo death or relapse event, an of risk the to respect with gain specified the without patients brvain:AL ct ypolsi ekei;C,cndneitra;ES vn-resria;HH ihhpriliy S vrl survi overall OS, hyperdiploidy; high HeH, survival; event-free EFS, interval; confidence CI, leukaemia; rate. relapse lymphoblastic RR, chromosome; Philadelphia acute Ph, ALL, Abbreviations: 2. Table and fec hoooewsasse niiulyaogal12P-eaieALptet hrfr,icmlt aytps( karyotypes incomplete Therefore, patient. ALL Ph-negative 112 all among individually assessed was chromosome each of gained) yfluorescence by ec h oa ubro ae nec o oe varied. model Cox each in cases of number total the Hence rtioy2 ( 20 trisomy or ) Chromosome 42 .1(.908) .204 02–.8,00 .0(.709) 0.02 (0.27–0.91), 0.2 0.50 (0.80–2.66), 1.46 0.05 (0.23–0.99), 0.48 0.5 (0.59–2.74), 1.28 0.001 0.2 0.05 (1.59–6.64), (0.37–1.19), (0.22–0.98), 3.25 0.67 0.47 0.4 (0.63–3.02), 0.08 1.38 (0.33–1.07), 0.59 0.06 0.9 (0.33–1.01), (0.30–3.26), 0.3 0.58 0.99 (0.72–2.91), 1.45 0.003 0.6 0.5 (1.62–9.94), (0.62–2.22), (0.54–3.78), 4.01 1.18 1.43 0.02 (0.29–0.89), 0.7 0.51 (0.41–1.81), 0.05 0.6 0.86 (0.32–1.00), (0.59–2.67), 0.56 1.25 0.2 (0.79–2.51), 0.2 0.4 1.41 (0.28–1.24), (0.37–1.52), 0.59 0.75 0.05 (0.25–1.00), 0.2 0.3 0.50 (0.30–1.26), (0.67–4.32), 0.6 0.62 1.70 0.001 (0.49–3.37), (1.55–6.39), 0.8 1.29 3.15 (0.59–2.00), 0.2 1.09 0.3 (0.18–1.50), (0.69–2.94), 0.52 1.42 0.05 0.3 (1.01–4.19), (0.43–1.31), 0.3 2.06 0.9 0.75 (0.32–1.36), (0.34–2.78), 0.66 0.05 0.97 (0.32–0.99), 0.5 0.2 0.56 (0.25–1.96), (0.21–1.43), 0.002 0.70 0.05 0.54 (0.23–0.72), (0.34–0.99), 0.41 0.2 0.5 0.58 (0.73–3.69), (0.06–3.48), 1.64 0.4 0.47 74/29 (0.67–2.70), 0.7 1.34 (0.41–1.78), 0.9 0.86 (0.55–1.91), 1.02 33/76 0.3 0.06 (0.66–4.59), (0.35–1.03), 1.74 0.8 96/14 0.60 (0.51–2.31), 1.09 0.01 0.9 0.4 13/86 (0.19–0.81), (0.28–3.11), (0.38–1.45), 0.39 0.94 0.74 0.8 0.5 17/82 (0.36–3.99), (0.53–3.41), 1.20 1.35 0.8 67/38 (0.61–1.93), 1.09 0.06 45/61 (0.96–3.93), 1.95 0.004 69/39 (0.25–0.76), 0.44 0.5 19/90 (0.24–1.88), 0.68 0.3 30/80 (0.70–3.50), 1.57 70/37 15/94 Y 29/71 X 10/95 22 65/37 21 16/83 20 70/35 19 12/87 18 13/86 17 14 12 11 10 9 8 7 6 5 4 2 1 c eut fuiait o ersinmdl oprn 1 e hngtv L ainswt atclrcrmsmlgi ihthose with gain chromosomal particular a with patients ALL Ph-negative HeH 112 comparing models regression Cox univariate of Results ucm fhg yedpod(16 hoooe)ALptet rae nteUALI/CG93til ( trial. UKALLXII/ECOG2993 the in treated patients ALL chromosomes) (51–65 hyperdiploid high of Outcome 3 copiesofchr4(+4) cd ab a nsitu in 2 copiesofchr4 d Number atrisk ubra ikNumberatrisk Number atrisk ). Ph-negative Ph-positive yrdsto ny( only hybridisation ubro ae with/without cases of Number Overall Survival (%) Relapse Rate (%) ano chromosome of gain 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0 47 85 54 34 41 45 52 58 71 84 107 70 35 5 8 2 1 7 4 2 644832262521161410765452822211714128654 135790012345678910 012345678910 012345678910 93 .7(.337) . .6(.426) . .8(.234) 0.3 (0.72–3.44), 1.58 0.9 (0.34–2.69), 0.96 0.1 (0.83–3.75), 1.77 19/34 15 94 93 82 018 20 22 28 35 39 42 49 56 61 31 41 099877 7 8 9 9 10 13 14 17 23 iesnedanss(er)Time sincediagnosis(years) Time sincediagnosis(years) Time sincediagnosis(years) ihhpriliyi dl ALL adult in hyperdiploidy High n ¼ a 1 eeicue nteaaye,i ufiin aawr vial o h hoooeudrconsideration. under chromosome the for available were data sufficient if analyses, the in included were 11) c aahv nybe rsne o hoooe htwr andi tlat1 patients. 10 least at in gained were that chromosomes for presented been only have Data and b d So hngtv ainssrtfidacrigt h rsneo bec ftioy4 trisomy of absence or presence the to according stratified patients Ph-negative of OS ) 3 copiesofchr4(+4) 2 copiesofchr4 Chilton L c ae only. Males 28 Ph-positive Ph-negative 624 26 tal et F ROS RR EFS 3 copiesofchr20(+20) 2 copiesofchr20 Number atrisk Ph-negative Ph-positive

aadrto(5 CI), (95% ratio Hazard Overall Survival (%) Overall Survival (%) 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 11288776657514638312826 67 55 94 03 82 23 25 28 35 40 43 49 136433322222 57 65 74 86 012345678910 Time sincediagnosis(years) P & -value 04McilnPbihr Limited Publishers Macmillan 2014 3 copiesofchr20(+20) 2 copiesofchr20 n ¼ 2 n ae identified cases and 12) Ph-positive Ph-negative b h tts(gain/not status The a and b Rand RR ) val; High hyperdiploidy in adult ALL L Chilton et al 1515 (P ¼ 0.2 and P ¼ 0.3, respectively). There was no evidence that the treatment with imatinib: 4.01 (95% CI: 1.74–9.25), P ¼ 0.001; 10.37 outcome of Ph-negative HeH patients correlated with increasing (95% CI: 2.75–39.17), P ¼ 0.001; and 3.73 (95% CI: 1.51–9.21), MCN. We examined the effect of each extra chromosome between P ¼ 0.004, respectively. Although only 13 Ph-positive patients 51 and 65 chromosomes by adding a categorical variable into a harboured þ 5, two failed to achieve a CR, seven relapsed and two Cox regression model. The resulting hazard ratio for risk of death died in first remission. Overall, 10/13 Ph-positive patients with þ 5 was 1.02 (95% CI: 0.93–1.12) (P ¼ 0.7). Similarly, no effect was died, including 6/7 of those patients who relapsed. The two observed when we grouped the cases into three subgroups patients with þ 5 who remained alive in first remission were both (51–53 vs 54–57 vs 58–65 chromosomes): 1.02 (95% CI: 0.63–1.65) male and aged 15 years. In addition, both patients received (P ¼ 0.9). imatinib and went on to have allogenic transplants. Although Both age and WCC are important risk factors in adult ALL. none of the other chromosomes were statistically significantly Therefore, we performed multivariate Cox regression modelling to associated with outcome (Table 5), it is noteworthy that the assess the prognostic relevance of all eight chromosomal gains in hazard ratios for Ph-positive patients with þ 4 were very similar to the context of these risk factors. Although only þ 4 and þ 20 those observed in the Ph-negative cohort (Tables 2 and 5). Only were shown to be independent markers of superior and inferior 3/48 Ph-positive patients gained chromosome 20 but all three outcome, respectively, they were significant for all three end patients died: one after failing to achieve CR and two after relapse. points (Table 3). Table 4 and Figures 2c and d show the The gain of chromosome 2 was significantly more prevalent demographic, clinical and outcome characteristics of Ph-negative among Ph-positive HeH patients compared with Ph-negative patients stratified by the presence or absence of these chromoso- patients. However, there was no evidence that this trisomy was mal gains. Previous analyses of the UKALLXII/ECOG2993 trial associated with an inferior outcome (Table 5). revealed that patients receiving an allograft transplant in first remission had improved outcome.11 The addition of a variable representing first remission treatment to the final Cox models (Table 3) did not significantly affect the direction or magnitude of DISCUSSION the hazard ratios for trisomy of chromosome 4 or 20. Younger This is the largest and most comprehensive study of HeH in adults (o25 years) are now routinely treated on protocols that are adolescent and adult ALL published to date. Moreover, all patients separate from those of adults (424 years) following evidence were treated on a single trial, UKALLXII/ECOG2993, allowing us to suggesting that their outcome significantly improves with paedia- compare the prognosis of patients with and without a Ph and tric-like therapy.16–18 Therefore, we re-ran the aforementioned Cox examine outcome heterogeneity by individual chromosomal gain. models restricting the analysis to patients o25 years of age and Overall, the frequency of HeH was 13% and interestingly this did then to those 425 years of age. These age-restricted models did not vary between Ph-positive and Ph-negative cases. Therefore, not materially affect the direction or magnitude of the hazard ratios although HeH is less common in adult ALL compared with for trisomy of chromosome 4 or 20, indicating that these results paediatric ALL, where it accounts for 35% of BCP-ALL patients, it is were applicable across the age spectrum of this cohort. one of the most prevalent cytogenetic subgroups in adult BCP- Among the Ph-positive patients we were able to investigate the ALL.1 Few previous studies have reported HeH as a specific effect of 13 chromosomes on outcome (Table 5). Only the gain of cytogenetic subgroup, but those that have report a frequency of was significantly associated with an adverse 5–10%.19–22 Hence, our incidence of 13% is broadly comparable outcome, and the hazard ratios for EFS, RR and OS remained with the results of these studies, especially given the differing age significant in multivariate analysis, adjusting for age, WCC and profiles of most adult cohorts and the correlation of HeH with

Table 3. Results of multivariate Cox regression modellinga for the risk of an event, relapse or death in 112 HeH Ph-negative ALL patients

Variableb Hazard ratio (95% CI) P-value Hazard ratio including first remission P-value treatment in the model (95% CI)

EFS þ 4 0.51 (0.30–0.88) 0.02 0.40 (0.22–0.73) 0.003 þ 20 2.58 (1.25–5.34) 0.01 3.61 (1.61–8.11) 0.002 WCC 1.26 (1.03–1.56) 0.03 1.37 (1.10–1.70) 0.005 Age (years) 1.03 (1.01–1.05) 0.001 1.03 (1.01–1.05) 0.01 First remission treatmentc 0.35 (0.16–0.74) 0.006 RR þ 4 0.43 (0.21–0.86) 0.02 0.29 (0.14–0.61) 0.001 þ 20 4.33 (1.67–11.19) 0.003 7.02 (2.56–19.28) o0.001 WCC 1.34 (1.02–1.76) 0.03 1.47 (1.09–1.97) 0.01 Age (years) 1.00 (0.97–1.04) 0.7 1.00 (0.97–1.03) 1.0 First remission treatmentc 0.08 (0.02–0.34) 0.001 Overall survival þ 4 0.47 (0.27–0.84) 0.01 0.36 (0.19–0.68) 0.002 þ 20 2.63 (1.25–5.54) 0.01 3.32 (1.45–7.53) 0.004 WCC 1.35 (1.091.67) 0.006 1.47 (1.17–1.84) 0.001 Age (years) 1.04 (1.02–1.06) o0.001 1.04 (1.01–1.06) 0.001 First remission treatmentc 0.42 (0.19–0.90) 0.03 Abbreviations: ALL, acute lymphoblastic leukaemia; EFS, event-free survival; HeH, high hyperdiploidy; Ph, Philadelphia chromosome; RR, relapse rate; WCC, white cell count. aMultivariate Cox regression modelling was performed using a forward-selection stepwise modelling process using a base model of age and WCC and the likelihood-ratio test to compare models. bBinary variables representing the gain or not gain of chromosomes 4, 5, 10, 14, 17, 20, 21 and X were considered in the modelling process along with a log-transformed for white cell count and age as a continuous variable. In order to generate mutually exclusive groups and maximise the number of cases available for inclusion in the models we inferred the trisomy status of cases which were unknown as ‘not trisomy’. cSibling donor allogenic transplant or matched unrelated donor versus chemotherapy and autologous transplant.

& 2014 Macmillan Publishers Limited Leukemia (2014) 1511 – 1518 High hyperdiploidy in adult ALL L Chilton et al 1516 younger age. Given that most current clinical trials on adult ALL have a lower age limit of 25 years, it is noteworthy that 50% of our 0.001 0.001 HeH patients were aged between 25 and 59 years. Therefore, the o o

ot trisomy’. results of this study are relevant to future adult clinical trials as

b well as to those incorporating teenagers and young adults. All the karyotypes in this study of 160 adolescents and adults had between 51 and 65 chromosomes and the average number of gained chromosomes was nine. The pattern of gain was non- 17 (8–52) random, with chromosomes 4, 6, 10, 14, 17, 18, 21 and X being the

-value most frequently gained. Thus, the cytogenetic profile of this P subgroup is very similar to that observed in paediatric ALL.6–9,23–25 0.001 48 (31–64)

o Therefore, we can assume that it is representative of the same biological subgroup. Although it is well established that HeH can 10 b coexist with t(9;22), the proportion of HeH harbouring a Ph in this study was 30% compared with o1% in a previous UK paediatric HeH cohort.7 This difference most likely reflects the higher incidence of t(9;22) in adult ALL generally, which is B25%. 83 (48–99) Although most cases with concurrent HeH/t(9;22) show both

Rates at 5 years (95% CI), abnormalities in the stem line/main clone, a few cases (n ¼ 3) The status (that is, trisomy or not trisomy) was unknown for

a showed only t(9;22) in the parent clone, suggesting that t(9;22) 0.001 58 (32–87) 0.002 23 (6–47) 0.001 42 (26–63) was the primary and HeH the secondary event. This observation is o o 10 ount. consistent with the literature. The only unique cytogenetic EFS RR OS b feature of HeH in adolescent and adult ALL was the gain of chromosome 2, which was observed in 30% of cases, thus much more frequently than in paediatric ALL (5%).7 However, it was strongly correlated with Ph-positive ALL where it was observed in two-thirds of cases. Trisomy 2 is rarely observed in non-HeH Ph-positive cases12 (and personal observations); hence, the

(%) selective advantage it provides must only be relevant in the N

Deaths, context of a HeH clone. We have previously reported that HeH patients treated in the UKALLXII/ECOG2993 trial have a superior outcome within a N Ph-negative cohort.13 In addition, we have shown that Ph-positive patients with HeH have marginally improved (%) first outcome when compared with Ph-positive patients without Deaths in 12 remission, HeH. These observations suggest that HeH patients are more sensitive to anti-leukaemic therapy compared with other ALL patients. This may be explained by the observation that HeH cells

(%) in vitro more readily accumulate higher levels of methotrexate N poly-glutamates.26 The hypothesis that HeH confers treatment susceptibility to the leukaemic clone is supported by the findings of this study, which show that after adjusting for the effects of age,

(%) Relapses, WCC and treatment the outcome of HeH patients in the UKALLXII/ N ECOG2993 trial was not adversely affected by the presence of the 20 survived beyond 3 years. CR, Ph chromosome. This is not an argument for altering þ the treatment of Ph-positive patients but rather recognition that /l)

9 the presence of one genetic feature can ameliorate the adverse

10 effect of another. WCC, Â median

( For the first time in adolescent and adult ALL, we have investigated the prognostic relevance of MCN and individual chromosomal gains in HeH. There was no evidence for a

13). In order to generate mutually exclusive subgroups and include as many cases as possible in the analysis, we inferred the status ofrelationship these cases as ‘n between MCN and outcome in this cohort. This is in Age, 8,9 ¼ (years) median contrast to some recent reports on paediatric HeH cohorts but n consistent with earlier studies.7,24 Univariate analysis revealed moderately strong evidence for outcome heterogeneity with M:F

ratio respect to the gain of eight chromosomes: 4, 5, 10, 14, 17, 20, 21 and X. Interestingly, all but one of these chromosomes (#20) have been implicated as prognostic factors in paediatric ALL.6–9,23,24 (%) Sex However, multivariate analysis adjusting for the effects of age, N WCC and other chromosomes revealed that just three

7) and chromosome 20 ( chromosomes were independent risk factors: þ 4 and þ 20 20 7 5:2 28 3 6 (86) 1 (17) 3 (33) 4 (57) 43 (10–73) 20 (3–80) 43 (10–73) ¼ among Ph-negative patients and þ 5 among Ph-positive patients. þ n a However, it should be noted that there were trends towards

20 statistical significance for þ 4 and þ 5 among Ph-positive and Demographic, clinical and outcome features of Ph-negative ALL patients with HeH according to the presence or absence of the trisomies of 4 and 20 þ 4 and 420 63 6 36:27Ph-negative 3:3 19 43 4 patients, 16 respectively. 61 (97) 6 (100) 15 (25) 5 (83) In 8paediatric (11) 1 (17) 21 (33) ALL, 62 6 (48–72) (100) trisomy 17 (1–52) 4 has 27 (17–41) 69 (56–79) þ þ þ 4 20 4/ YesNoYes 70 42 41:29 13 28:14 19 28 8:5 4 33 5.5 4 67 (96) 40 (95) 16 (24) 16 (40) 12 (92) 9 (13) 8 6 (19) (50) 25 (36) 26 (62) 3 (23) 60 (47–70) 36 (22–51) 10 (77) 0.003 27 (17–40) 23 (6–47) 49 (33–68) 0.007 66 (54–76) 0.001 41 (26–56) No 99 61:38 20 4 95 (96) 26 (27) 14 (14) 41 (41) 55 (44–64) 32 (23–43) 62 (51–70) Neither 36 25:11 25 5 34 (94) 11 (32) 7 (21) 20 (56) 43 (26–58) Subgroup been reported to be associated with a favourable outcome by þ þ þ 6–9

Total 112 69:43 21 4 107 (96)several 32 (30) groups 17 (16)and 51 (46) is 51used (41–60) by the Children’s 35 (26–45) Oncology 57 (47–66) Group Rates are quoted at 2 years because none of the six patients with chromosome 4 ( Table 4. Abbreviations: CI, confidence interval; CR, complete remission;b EFS, event-free survival; OS, overall survival; RR, relapse rate; WCC, white cell c along with þ 10 and/or þ 17 to identify a subset of children with

Leukemia (2014) 1511 – 1518 & 2014 Macmillan Publishers Limited High hyperdiploidy in adult ALL L Chilton et al 1517 Table 5. Results of Cox regression models comparing the risk of an event, relapse or death in 48 HeH Ph-positive ALL patients with a specified chromosomal gain to those patients without the specified gain

Hazard ratio (95% CI)

Chromosomea Number of cases with/without EFS RR OS gain of chromosome

2 31/17 1.80 (0.82–3.94), 0.1 1.09 (0.38–3.19), 0.9 1.88 (0.79–4.49), 0.2 4 31/17 0.50 (0.24–1.02), 0.06 0.38 (0.13–1.11), 0.08 0.47 (0.22–1.04), 0.06 5 13/35 3.20 (1.45–7.08), 0.004 7.40 (2.26–24.28), 0.001 3.31 (1.42–7.72), 0.005 6 29/19 1.18 (0.55–2.51), 0.7 0.71 (0.24–2.06), 0.5 1.17 (0.51–2.68), 0.7 8 20/28 1.50 (0.73–3.09), 0.3 1.76 (0.62–5.03), 0.3 1.40 (0.65–3.03), 0.4 10 19/29 1.45 (0.70–3.00), 0.3 2.29 (0.78–6.72), 0.1 1.31 (0.60–2.86), 0.5 14 32/16 0.51 (0.24–1.04), 0.07 0.36 (0.12–1.03), 0.06 0.50 (0.23–1.10), 0.09 15 12/36 1.20 (0.53–2.71), 0.7 1.44 (0.44–4.71), 0.6 1.54 (0.66–3.59), 0.3 18 18/30 2.19 (1.06–4.56), 0.04 2.75 (0.94–8.06), 0.07 1.86 (0.85–4.06), 0.1 21 43/5 1.19 (0.36–3.93), 0.8 Nb 0.93 (0.28–3.11), 0.9 þ der(22)t(9;22) 38/10 1.48 (0.56–3.91), 0.4 1.15 (0.31–4.22), 0.8 1.18 (0.44–3.19), 0.7 X 33/15 0.86 (0.40–1.84), 0.7 0.54 (0.18–1.57), 0.3 0.65 (0.29–1.44), 0.3 Y 10/14 0.57 (0.17–1.87), 0.4 Nc 1.32 (0.35–4.92), 0.7 Abbreviations: CI, confidence interval; EFS, event-free survival; OS, overall survival; RR, relapse rate. Cox regression models included a variable for treatment with imatinib. aData has only been presented for chromosomes that were gained in at least 10 patients. bNone of the patients without gain of chromosome 21 relapsed therefore estimate of hazard ratio is not possible. cNone of the patients with gain of chromosome Y relapsed therefore estimate of hazard ratio is not possible. very low-risk ALL.27 However, our analysis suggests that in adult from that seen in paediatric ALL. The exception is the frequency of ALL the addition of þ 10 or þ 17 does not confer any additional cases with concurrent HeH/t(9;22) and the related higher advantage over and above that exerted by þ 4. Also, þ 5 has incidence of trisomy of chromosome 2. In addition, we have been previously associated with an adverse outcome by Children’s demonstrated significant outcome heterogeneity within the HeH Oncology Group.24 In contrast, the adverse prognostic effect of subgroup in relation to the presence of specific individual þ 20 is a novel finding. The gain of chromosome 20 is less chromosomal gains. Given the recent remodelling of adult ALL frequent among children with HeH (6%); thus, its adverse effect treatment protocols in terms of age range and TKI-based therapy may have been masked. However, in this adolescent and adult for Ph-positive patients, our findings will need to be replicated in cohort those patients with þ 20 had an extremely poor outcome, more contemporary protocols before their clinical relevance can which was largely driven by a high rate of early relapse. be accurately assessed. However, these data argue strongly for the The functional consequence of the simultaneous gain of prospective cytogenetic characterisation of all adult ALL patients. multiple chromosomes is not fully understood, but one plausible Only by doing this consistently and accurately, we will be able to hypothesis is that it results in overexpression of one or more investigate HeH as a discrete and meaningful biological subgroup genes on each of the gained chromosomes.28 This idea is in the future. supported by in vitro studies showing that the accumulation of methotrexate poly-glutamates, a derivative of methotrexate, is greater in the blasts of HeH patients compared with non-HeH CONFLICT OF INTEREST patients.26 This phenomenon is thought to explain, in part, the The authors declare no conflict of interest. good outcome of paediatric HeH patients. The reason why HeH cells accumulate higher levels of methotrexate poly-glutamates could be the presence of an extra copy of the reduced folate ACKNOWLEDGEMENTS carrier gene that governs the uptake of methotrexate into the We thank the following: (1) the UK Cancer Cytogenetics Group (UKCCG) and the cell.26 The reduced folate carrier gene is located on chromosome Eastern Co-operative Oncology Group (ECOG) cytogenetic laboratories for providing 21, which is almost ubiquitously gained in paediatric HeH ALL and data; (2) Gordon Dewald and past and present members of the Leukaemia Research was gained in 90% of the cases in this study. Therefore, Cytogenetics Group for collecting and reviewing the cytogenetic data; and (3) the extrapolating this logic, it seems plausible that the gene dosage doctors who participated in the UKALLXII/ECOG2993 trial. AVM and CJH also thank Leukaemia & Research (LLR) for financial support. We also thank Dr Sue effect produced by trisomies 5/20 and 4 is more or less likely to Richards (Clinical Trial Service Unit, University of Oxford) for data management and confer drug resistant properties on the leukaemic clone. statistical input. Interestingly þ 20 has been associated with inferior outcome in Ewing’s Sarcoma where it is also usually observed in the context of multiple chromosomal gain.29 Further research is required to AUTHOR CONTRIBUTIONS unravel the molecular consequence of HeH in all . AVM and LC designed the study, performed statistical analyses and wrote the Identifying individual trisomies that confer prognostic relevance, manuscript; GB provided trial data and statistical advice; AVM, CJH and RPK especially across multiple diseases, could provide important clues classified the cytogenetic abnormalities and managed the UK and ECOG for targeting such research towards those genes that are most cytogenetic databases; JMR, MST, AHG and AKF co-ordinated the clinical trial; likely to be involved in treatment response. and all authors approved the final version of the manuscript. In conclusion, this study has clearly demonstrated that HeH is an important cytogenetic subtype of adolescent and adult ALL. It is present in equal frequencies in Ph-positive and Ph-negative REFERENCES patients and is not confined solely to younger adults because 50% 1 Moorman AV. The clinical relevance of chromosomal and genomic abnormalities of our patients were aged 25 years or older. The cytogenetic in B-cell precursor acute lymphoblastic leukaemia. Blood Rev 2012; 26: profile of HeH in this age group was not significantly different 123–135.

& 2014 Macmillan Publishers Limited Leukemia (2014) 1511 – 1518 High hyperdiploidy in adult ALL L Chilton et al 1518 2 Maia AT, Tussiwand R, Cazzaniga G, Rebulla P, Colman S, Biondi A et al. 15 Harrison CJ, Moorman AV, Barber KE, Broadfield ZJ, Cheung KL, Harris RL et al. Identification of preleukemic precursors of hyperdiploid acute lymphoblastic Interphase molecular cytogenetic screening for chromosomal abnormalities of leukemia in cord blood. Genes Chromosomes Cancer 2004; 40: 38–43. prognostic significance in childhood acute lymphoblastic leukaemia: a UK Cancer 3 Davidsson J, Paulsson K, Lindgren D, Lilljebjorn H, Chaplin T, Forestier E et al. Cytogenetics Group Study. Br J Haematol 2005; 129: 520–530. Relapsed childhood high hyperdiploid acute lymphoblastic leukemia: presence of 16 Storring JM, Minden MD, Kao S, Gupta V, Schuh AC, Schimmer AD et al. Treatment preleukemic ancestral clones and the secondary nature of microdeletions and of adults with BCR-ABL negative acute lymphoblastic leukaemia with a modified RTK-RAS . Leukemia 2010; 24: 924–931. paediatric regimen. Br J Haematol 2009; 146: 76–85. 4 Paulsson K, Forestier E, Lilljebjorn H, Heldrup J, Behrendtz M, Young BD et al. 17 Huguet F, Leguay T, Raffoux E, Thomas X, Beldjord K, Delabesse E et al. Genetic landscape of high hyperdiploid childhood acute lymphoblastic leukemia. Pediatric-inspired therapy in adults with Philadelphia chromosome-negative acute Proc Natl Acad Sci USA 2010; 107: 21719–21724. lymphoblastic leukemia: the GRAALL-2003 study. J Clin Oncol 2009; 27: 911–918. 5 Paulsson K, Johansson B. High hyperdiploid childhood acute lymphoblastic 18 Ramanujachar R, Richards S, Hann I, Goldstone A, Mitchell C, Vora A et al. leukemia. Genes Chromosomes Cancer 2009; 48: 637–660. Adolescents with acute lymphoblastic leukaemia: outcome on UK national 6 Sutcliffe MJ, Shuster JJ, Sather HN, Camitta BM, Pullen J, Schultz KR et al. paediatric (ALL97) and adult (UKALLXII/E2993) trials. Pediatr Blood Cancer 2007; High concordance from independent studies by the Children’s Cancer Group 48: 254–261. (CCG) and Pediatric Oncology Group (POG) associating favorable prognosis with 19 GFCH. Cytogenetic abnormalities in adult acute lymphoblastic leukemia: combined trisomies 4, 10, and 17 in children with NCI Standard-Risk B-precursor correlations with hematologic findings and outcome. A Collaborative Study of the Acute Lymphoblastic Leukemia: a Children’s Oncology Group (COG) initiative. Groupe Francais De Cytogenetique Hematologique. Blood 1996; 87: 3135–3142. Leukemia 2005; 19: 734–740. 20 Martin Ramos ML, Lahuerta Palacios JJ, Martinez LJ, Gomez Rodriguez MJ, 7 Moorman AV, Richards SM, Martineau M, Cheung KL, Robinson HM, Jalali GR et al. Moreno IA, Barreiro ME. Karyotype and prognosis in adult Spanish acute Outcome heterogeneity in childhood high-hyperdiploid acute lymphoblastic lymphoblastic leukemia. Haematologica 2001; 86: 438–439. leukemia. Blood 2003; 102: 2756–2762. 21 Mi JQ, Wang X, Yao Y, Lu HJ, Jiang XX, Zhou JF et al. Newly diagnosed acute 8 Dastugue N, Suciu S, Plat G, Speleman F, Cave H, Girard S et al. Hyperdiploidy lymphoblastic leukemia in China (II): prognosis related to genetic abnormalities in with 58-66 chromosomes in childhood B-acute lymphoblastic leukemia is highly a series of 1091 cases. Leukemia 2012; 26: 1507–1516. curable: 58951 CLG-EORTC results. Blood 2013; 121: 2415–2423. 22 Gomez-Segui I, Cervera J, Such E, Martinez-Cuadron D, Luna I, Ibanez M et al. 9 Paulsson K, Forestier E, Andersen MK, Autio K, Barbany G, Borgstrom G et al. Prognostic value of cytogenetics in adult patients with Philadelphia-negative High modal number and triple trisomies are highly correlated favorable factors in acute lymphoblastic leukemia. Ann Hematol 2012; 91: 19–25. childhood B-cell precursor high hyperdiploid acute lymphoblastic leukemia 23 Heerema NA, Raimondi SC, Anderson JR, Biegel J, Camitta BM, Cooley LD et al. treated according to the NOPHO ALL 1992/2000 protocols. Haematologica 2013; Specific extra chromosomes occur in a modal number dependent pattern in 98: 1424–1432. pediatric acute lymphoblastic leukemia. Genes Chromosomes Cancer 2007; 46: 10 Paulsson K, Harrison CJ, Andersen MK, Chilton L, Nordgren A, Moorman AV et al. 684–693. Distinct patterns of gained chromosomes in high hyperdiploid acute lympho- blastic leukemia with t(1;19)(q23;p13), t(9;22)(q34;q22) or MLL rearrangements. 24 Heerema NA, Sather HN, Sensel MG, Zhang T, Hutchinson RJ, Nachman JB et al. Leukemia 2013; 27: 974–977. Prognostic impact of trisomies of chromosomes 10, 17 and 5 among children with 11 Goldstone AH, Richards SM, Lazarus HM, Tallman MS, Buck G, Fielding AK et al. acute lymphoblastic leukemia and high hyperdiploidy (450 chromosomes). J Clin In adults with standard-risk acute lymphoblastic leukemia, the greatest benefit is Oncol 2000; 18: 1876–1887. achieved from a matched sibling allogeneic transplantation in first complete 25 Raimondi SC, Pui CH, Hancock ML, Behm FG, Filatov L, Rivera GK. Heterogeneity of remission, and an autologous transplantation is less effective than conventional hyperdiploid (51-67) childhood acute lymphoblastic leukemia. Leukemia 1996; 10: consolidation/maintenance chemotherapy in all patients: final results of the 213–224. International ALL Trial (MRC UKALL XII/ECOG E2993). Blood 2008; 111: 1827–1833. 26 Whitehead VM, Vuchich MJ, Lauer SJ, Mahoney D, Carroll AJ, Shuster JJ et al. 12 Fielding AK, Rowe JM, Richards SM, Buck G, Moorman AV, Durrant IJ et al. Accumulation of high levels of methotrexate polyglutamates in lymphoblasts Prospective outcome data on 267 unselected patients with Philadelphia- from children with hyperdiploid (450 chromosomes) B-lineage acute lympho- chromosome positive adult acute lymphoblastic leukaemia confirms superiority blastic leukemia: a Pediatric Oncology Group study. Blood 1992; 80: 1316–1323. of allogeneic transplant over chemotherapy the pre-imatinib era Results from 27 Chauvenet AR, Martin PL, Devidas M, Linda SB, Bell BA, Kurtzberg J et al. Anti- the international ALL trial MRC UKALLXII/ECOG2993. Blood 2009; 113: metabolite therapy for lesser-risk B-lineage acute lymphoblastic leukemia of 4489–4496. childhood: a report from Children’s Oncology Group Study P9201. Blood 2007; 13 Moorman AV, Harrison CJ, Buck GA, Richards SM, Secker-Walker LM, Martineau M 110: 1105–1111. et al. Karyotype is an independent prognostic factor in adult acute lymphoblastic 28 Gruszka-Westwood AM, Horsley SW, Martinez-Ramirez A, Harrison CJ, Kempski H, leukemia (ALL): Analysis of cytogenetic data from patients treated on the Medical Moorman AV et al. Comparative expressed sequence hybridization studies of Research Council (MRC) UKALLXII / Eastern Cooperative Oncology Group (ECOG) high-hyperdiploid childhood acute lymphoblastic leukemia. Genes Chromosomes 2993 Trial. Blood 2007; 109: 3189–3197. Cancer 2004; 41: 191–202. 14 Fielding AK, Rowe JM, Buck G, Foroni L, Gerrard G, Litzow MR et al. 29 Roberts P, Burchill SA, Brownhill S, Cullinane CJ, Johnston C, Griffiths MJ et al. UKALLXII/ECOG2993: addition of imatinib to a standard treatment regimen Ploidy and karyotype complexity are powerful prognostic indicators in the enhances long-term outcomes in Philadelphia positive acute lymphoblastic Ewing’s sarcoma family of tumors: a study by the United Kingdom Cancer leukemia. Blood 2013; e-pub ahead of print 25 November 2013; doi:10.1182/ Cytogenetics and the Children’s Cancer and Leukaemia Group. Genes Chromosomes blood-2013-09-529008. Cancer 2008; 47: 207–220.

Leukemia (2014) 1511 – 1518 & 2014 Macmillan Publishers Limited