VOLUME
32
䡠
NUMBER
14
䡠
MAY
10
2014
JOURNAL OF CLINICAL ONCOLOGY
O R I G I N A L
R E P O R T
IGH@ Translocations Are Prevalent in Teenagers and Young Adults With Acute Lymphoblastic Leukemia and Are Associated With a Poor Outcome Lisa J. Russell, Amir Enshaei, Lisa Jones, Amy Erhorn, Dino Masic, Helen Bentley, Karl S. Laczko, Adele K. Fielding, Anthony H. Goldstone, Nicholas Goulden, Christopher D. Mitchell, Rachel Wade, Ajay Vora, Anthony V. Moorman, and Christine J. Harrison Lisa J. Russell, Amir Enshaei, Lisa Jones, Amy Erhorn, Dino Masic, Helen Bentley, Anthony V. Moorman, and Christine J. Harrison, Leukaemia Research Cytogenetics Group, Northern Institute for Cancer Research, Newcastle University, Newcastle-uponTyne; Karl S. Laczko, Leica Microsystems, Gateshead; Adele K. Fielding, Royal Free and University College London Medical School; Anthony H. Goldstone, University College London Hospital; Nicholas Goulden, Great Ormond St Hospital, London; Christopher D. Mitchell, John Radcliffe Hospital; Rachel Wade, Clinical Trial Service Unit, University of Oxford, Oxford; and Ajay Vora, Sheffield Children’s Hospital, Sheffield, United Kingdom. Published online ahead of print at www.jco.org on April 7, 2014. Supported by the Kay Kendall Leukemia Fund and Leukemia & Lymphoma Research (formerly Leukemia Research). A.V.M. and C.J.H. share senior authorship. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article. Corresponding author: Lisa J. Russell, PhD, Leukaemia Research Cytogenetics Group, Northern Institute for Cancer Research, Level 5, Sir James Spence Institute, Royal Victoria Infirmary, Newcastle-upon-Tyne, NE1 4LP, United Kingdom; e-mail: lisa.russell@ newcastle.ac.uk. © 2014 by American Society of Clinical Oncology 0732-183X/14/3214w-1453w/$20.00 DOI: 10.1200/JCO.2013.51.3242
A
B
S
T
R
A
C
T
Purpose To determine the prevalence and prognostic association of immunoglobulin heavy chain (IGH@) translocations in acute lymphoblastic leukemia (ALL). Patients and Methods The cohort comprised 3,269 patients treated on either the UKALL2003 trial for children and adolescents (1 to 24 years old) or the UKALLXII trial for adolescents and adults (15 to 59 years old). High-throughput fluorescent in situ hybridization was used to detect IGH@ translocations. Results We identified IGH@ translocations in 5% of patients with ALL (159 of 3,269 patients), in patients with both B-cell (148 of 2,863 patients) and T-cell (11 of 408 patients) disease. Multiple partner genes were identified including CRLF2 (n ⫽ 35), five members of the CEPB gene family (n ⫽ 17), and ID4 (n ⫽ 11). The level of the IGH@-positive clone varied and indicated that some IGH@ translocations were primary events, whereas others were secondary aberrations often associated with other established aberrations. The age profile of patients with IGH@ translocations was distinctive, with a median age of 16 years and peak incidence of 11% among 20- to 24-year-old patients. Among patients with B-cell precursor ALL who were Philadelphia chromosome negative, those with an IGH@ translocation had an inferior overall survival compared with other patients (UKALL2003: hazard ratio, 2.37; 95% CI, 1.34 to 4.18; P ⫽ .003; UKALLXII: hazard ratio, 1.73; 95% CI, 1.22 to 2.47; P ⫽ .002). However, this adverse effect was not independent of age or minimal residual disease status and did not seem to be driven by an increased risk of relapse. Conclusion IGH@ translocations define a genetic feature that is frequent among adolescents and young adults with ALL. Although associated with an adverse outcome in adults, it is not an independent prognostic factor in children and adolescents. J Clin Oncol 32:1453-1462. © 2014 by American Society of Clinical Oncology
INTRODUCTION
Genetic aberrations are important indicators of prognosis in acute lymphoblastic leukemia (ALL) and are widely used in risk stratification.1 Translocations involving the immunoglobulin heavy chain locus (IGH@) are hallmarks of mature B-cell malignancies, where they drive pathogenesis (reviewed in Molyneux et al2 and Freedman3). IGH@ translocations have been described in B-cell precursor ALL (BCP-ALL), where they target different genes with the same consequence; the partner gene is overexpressed as a result of its close proximity to the IGH@ enhancer. Previously reported recurrent BCP-ALL translocation partner genes include five members of
the CCAAT/enhancer binding protein (CEBP) family of transcription factors, showing opposing functions for deregulation in myeloid and lymphoid leukemogenesis4,5; the inhibitory transcription factor, ID4, in the translocation, t(6;14)(p22;q32), defining a subgroup characterized by deletion of CDKN2A/B and PAX56; the cytokine receptor for erythropoietin (EPOR) at 19p13; type I cytokine receptor, CRLF2, at Xp22/Yp117,8; and other sporadic partner genes.9-12 We recently reported the incidence and clinical relevance of IGH@ translocations in Philadelphia chromosome–negative adult BCPALL.13 However, the incidence, spectrum, and prognostic relevance of IGH@ translocations in © 2014 by American Society of Clinical Oncology
Information downloaded from jco.ascopubs.org and provided by at University of Brighton on July 10, 2014 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 194.81.203.94
1453
Russell et al
ALL across all age groups remained unknown. In this study, we developed a high-throughput fluorescent in situ hybridization (FISH) screening approach to ascertain the incidence of these translocations among 3,269 patients and examined their prognostic relevance.
PATIENTS AND METHODS Patients Between October 2003 and June 2011, 3,182 children and adolescents, age 1 to 24 years at diagnosis, were treated on the UKALL2003 trial. Between January 1993 and November 2008, 1,380 adolescents and adults, age 15 to 55 years, were treated on UKALLXII. Together, 4,562 patients were eligible for this study (Fig 1). Clinical details, including age, sex, WBC count, immunophenotype, and survival data, were collected by the Clinical Trial Service Unit, University of Oxford (Oxford, United Kingdom). Institutional review board approval was obtained at each of the collaborating centers. Informed consent was obtained in accordance with the Declaration of Helsinki. In UKALL2003, minimal residual disease (MRD) was evaluated by realtime quantitative polymerase chain reaction analysis of immunoglobulin and T-cell receptor gene rearrangements as defined by the European MRD Study
Group.14 Patients were classified by MRD at day 28 (end of induction) as high (⬎ 0.01%) or low (⬍ 0.01%) risk. Cytogenetics and FISH Fixed cells were available from 3,302 patients, including 2,595 children (2,279 with BCP-ALL, 54 with Down syndrome and ALL [DS-ALL], and 316 with T-cell ALL [T-ALL]) and 707 adults (615 with BCP-ALL and 92 with T-ALL; Fig 1). Patients were screened for IGH@ translocations using a high-throughput automated capture and scoring FISH approach or FISH by eye (Appendix, online only). The involvement of IGH@ was determined using a variety of IGH@ FISH probes (Appendix Table A1, online only). Images were captured and scored using Cytovision SPOT counting software (Leica Microsystems, Gateshead, United Kingdom). Because IGH@-CRLF2 is cryptic, FISH for rearrangements involving CRLF2 was performed on patients with an IGH@-positive FISH result.8 Patients with known, cytogenetically visible translocations were tested with appropriate FISH probes, as previously reported.4-8,12 Multiplex Ligation-Dependent Probe Amplification DNA extracted directly from the bone marrow aspirate or fixed cells used for FISH was available from 58 UKALL2003 and 30 UKALLXII IGH@-positive patients with BCP-ALL. Copy number alterations were investigated by multiplex ligation-dependent probe amplification (MLPA) using the SALSA MLPA kit P335-A1 (MRC Holland, Amsterdam, the Netherlands).15 The kit includes probes to IKZF1, CDKN2A/B, PAX5, EBF1, ETV6, BTG1, RB1, and CSF2RA/IL3RA/CRLF2.
Total patients (N = 4,562) Excluded FISH not tested
(n =1,260)
Available for incidence analysis (n = 3,302)
UKALL2003 trial IGH@ negative IGH@ positive
(n = 2,595) (n = 2,469) (n = 103)
UKALLXII trial IGH@ negative IGH@ positive
(n = 707) (n = 641) (n = 56) Excluded (n = 2,179) IGH@- and IGH status not known, MLPA not tested
Excluded (n = 2,179) IGH@- and IGH status not known, MLPA not tested MLPA analysis (n = 58)
MLPA analysis (n = 30)
Excluded (n = 258) Lost to follow-up, t(9;22) and T cell
Excluded (n = 406) Lost to follow-up, t(9;22) and T cell Survival analysis (n = 2,194)
Survival analysis (n = 449)
Fig 1. Analysis profile for investigation of the incidence of IGH@ translocations, survival analysis, and prevalence of specific gene deletions. A total of 4,562 patients were entered onto two treatment trials, the UKALL2003 and UKALLXII trials. Fixed material was available from 3,302 patients, including 2,595 childhood patients from UKALL2003 and 707 adults from UKALLXII; the remaining 1,260 patients without fixed material were excluded. Before multiplex ligation-dependent probe amplification (MLPA) or survival analysis, patients who did not have an IGH@ translocation, had T-cell disease, or had Philadelphia chromosome–positive acute lymphoblastic leukemia were excluded from further analysis. From a total of 103 childhood and 56 adult IGH@-positive patients, DNA was available from 58 and 30 of them, respectively, for MLPA analysis. Survival data were available from 2,194 children and 449 adults who had a known IGH@ result, both positive and negative. Patients with no follow-up data were removed before analysis. FISH, fluorescent in situ hybridization. 1454
© 2014 by American Society of Clinical Oncology
JOURNAL OF CLINICAL ONCOLOGY
Information downloaded from jco.ascopubs.org and provided by at University of Brighton on July 10, 2014 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 194.81.203.94
IGH@ Translocations in ALL
Table 1. Demographics and Clinical Characteristics of UKALL2003 Patients With and Without IGH@ Translocations Demographic or Clinical Characteristic Total Sex Female Male Age, years 1-4 5-9 10-14 15-19 20-24 Median WBC count, ⫻ 109/L ⬍ 50 ⬎ 50 Median NCI risk groupⴱ Standard risk High risk Immunophenotype B cell T cell DS-ALL No Yes MRD risk group† High Low Ph status Negative Positive
Total Patients
IGH@-Negative Patients
Patients With Failed Tests
IGH@-Positive Patients
No.
%
No.
% of Total
No.
% of Total
2,595
100
2,469
95
103
4
P
No.
% of Total
23
1
8 15
0 0
13 7 1 1 1
1 1 0 0 2
.42 1,112 1,483
100 100
1,064 1,405
96 95
40 63
4 4
1,246 649 430 221 49
100 100 100 100 100
1,212 622 402 194 39
97 96 93 88 80
21 20 27 26 9
2 3 6 12 18
⬍ .001
5
5
12
5 .91
1,991 604
100 100
1,892 577
13
95 96
80 23
14
4 4
19 4
12
0 0 7
⬍ .001 1,486 1,109
100 100
1,437 1,032
97 93
33 70
2 6
2,279 316
100 100
2,162 307
95 97
95 8
4 3
2,538 57
100 100
2,423 46
95 81
94 9
4 16
911 849
100 100
880 790
97 93
21 53
2 6
2,541 54
100 100
2,417 52
95 96
101 2
4 4
16 7
1 1
22 1
0 0
21 2
1 3
10 6
1 1
23 0
1 0
.22
.02
⬍ .001
1.0
Abbreviations: DS-ALL, Down syndrome and acute lymphoblastic leukemia; MRD, minimal residual disease; NCI, National Cancer Institute; Ph, Philadelphia translocation. ⴱ NCI risk stratifies risk according to age and WBC count. Standard risk indicates WBC count less than 50,000/mL and age 1 to younger than 10 years. High risk indicates WBC count ⱖ 50,000/mL and age ⱖ 10 years. †MRD was evaluated by real-time quantitative polymerase chain reaction analysis of immunoglobulin and T-cell receptor gene rearrangements as defined by the European MRD Study Group. Patients were stratified by MRD at the end of induction into high-risk (⬎ 0.01%) and low-risk (⬍ 0.01%) groups. MRD risk group could not be assigned for 728 patients. Data on MRD are not available for 837 patients.
Survival Survival analysis was restricted to patients with BCP-ALL, and separate analyses were performed for UKALL2003 and UKALLXII. Philadelphia chromosome–positive patients were excluded from analysis, because many received imatinib. Therefore, 2,194 children and young adults (UKALL2003) and 449 adults (UKALLXII) were analyzed (Fig 1). Survival analysis considered the following three end points: event-free survival, defined as time to relapse or death or second neoplasm, with censoring at last contact; relapse rate, defined as time from complete remission (CR) to relapse, with censoring at death in remission or last contact; and overall survival, defined as time to death, with censoring at last contact. Patients who did not achieve a CR were included in the analysis until an event or censor. Survival rates were calculated and compared using the Kaplan-Meier method, log-rank test, and Cox regression models (univariate and multivariable analyses). The Fisher’s exact test was used to examine the distribution of categorical variables. Because of the investigative nature of this analysis, we did not apply stringent multiple comparisons adjustment. However, all tests were conducted at the 1% significance level. All analyses were performed using Intercooled Stata 11.0 for Windows (Stata Corporation, College Station, TX). www.jco.org
RESULTS
Genetic Characterization of IGH@ Translocations Incidence of IGH@ translocations and the partner genes. IGH@ FISH was successful for 3,269 patients; 33 patients (UKALL2003, n ⫽ 23; UKALLXII, n ⫽ 10) had two failed independent tests (Tables 1 and 2). Patients who were tested did not differ from those not tested with respect to sex, age, WBC count, and National Cancer Institute (NCI) risk group (Appendix Tables A2 and A3, online only). However, in the adult cohort, fewer patients with T-ALL were tested (P ⬍ .002; Appendix Table A3). IGH@ translocations were observed in 5% of patients (159 of 3,269 patients; Tables 1 and 2), including 148 (5%) of 2,863 patients with BCP-ALL and 11 (3%) of 407 patients with T-ALL. CRLF2 was the most prevalent IGH@ translocation partner gene, identified in 35 (22%) of 159 IGH@-positive patients but not in patients with T-ALL (Appendix Table A4, online only; Fig 2A). The © 2014 by American Society of Clinical Oncology
Information downloaded from jco.ascopubs.org and provided by at University of Brighton on July 10, 2014 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 194.81.203.94
1455
Russell et al
Table 2. Demographics and Clinical Characteristics for UKALLXII Adult Patients With and Without IGH@ Translocations Demographic or Clinical Characteristic Total Sex Female Male WBC count, ⫻ 109/L ⬍ 50 ⬎ 50 Median Age, years 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60⫹ Median Immunophenotype B cell T cell Ph status Negative Positive
IGH@-Negative Patients
Total Patients
Patients With Failed Tests
IGH@-Positive Patients
No.
%
No.
% of Total
No.
% of Total
707
100
641
91
56
8
P
No.
% of Total
10
1
4 6
1 1
8 2
1 1
.57 288 419
100 100
259 382
90 91
25 31
9 7
518 180
100 100
471 162
91 90
39 16
8 9
.63
14
14
15
7 .98
119 95 79 60 60 74 69 54 21 10
100 100 100 100 100 100 100 100 100 100
119 95 79 60 60 74 69 54 21 10
33
89 91 93 90 91 94 87 92 91 83
12 8 5 6 5 4 9 3 2 2
33
9 8 6 9 8 5 11 5 9 17
2 1 1 1 1 1 1 2 0 0
33
2 1 1 1 1 1 1 2 0 0 33
.1 615 92
100 100
553 88
90 96
53 3
9 3
554 153
100 100
494 147
89 96
52 4
9 3
9 1
1 1
8 2
1 1
⬍ .001
Abbreviation: Ph, Philadelphia translocation.
CEBP gene family and ID4 were involved in 17 (11%) of 159 and 11 (7%) of 159 IGH@-positive BCP-ALLs, respectively (Fig 2A), and were also not identified in T-ALL. BCL2 was involved in four patients, and other partner genes were rare (ⱕ two patients each) (Appendix Table A4, Fig 2A). In T-ALL, one novel partner was identified (TAL1). The involvement of TRA/D@ was observed in both B- and T-cell disease. The percentage of IGH@-positive nuclei ranged from 5% to 100% and varied according to the partner gene (Fig 2B). For some partners (BCL2, CEBPA, CEBPB, and CRLF2), rearrangements were present in the majority of nuclei (median, 79%), whereas the proportion was lower for other genes (ID4: median, 65%; CEPBE: median, 30%; and unknown partner genes: median, 30%). Cytogenetic characterization of IGH@-positive patients. Among IGH@-positive patients (n ⫽ 159), 37% had a normal karyotype (n ⫽ 43) or a failed cytogenetics test (n ⫽ 16). The IGH@ translocation was the sole abnormality in 11 patients, whereas additional cytogenetic changes were present in the majority of patients (89 of 100 patients; Fig 3; Appendix Table A4). IGH@ translocations were observed with an established cytogenetic aberration in 28 (19%) of 148 patients with BCP-ALL, including BCR-ABL1 (n ⫽ 6), intrachromosomal amplification of chromosome 21 (iAMP21; n ⫽ 3), MLL rearrangements (n ⫽ 2), high hyperdiploidy (n ⫽ 9), and ETV6-RUNX1 (n ⫽ 8). The partner gene was known in only six of these patients (Appendix Table A4). In T-ALL, single IGH@-positive patients had rearrangements of STIL-TAL1, 1456
© 2014 by American Society of Clinical Oncology
MLL, TAL2, MLLT10-PICALM, LMO2, HOX11, BCL11B, TLX3, TRB@, and TRA/D@ (Appendix Table A4). Copy number abnormalities. We assessed the copy number status of eight genes (IKZF1, CDKN2A/B, PAX5, EBF1, ETV6, BTG1, and RB1) and the presence of P2RY8-CRLF2 fusion in 88 IGH@-positive patients with BCP-ALL using MLPA (UKALL2003, n ⫽ 58; UKALLXII, n ⫽ 30). There were no differences in sex, age, and WBC count between IGH@-positive patients tested and not tested by MLPA (Appendix Table A5, online only). Deletions were identified in all eight genes, with CDKN2A/B and IKZF1 deletions being the most prevalent (approximately 40%; Appendix Table A6, online only). Deletions were rarer in children age 1 to 9 years (average, 0.84 per patient) compared with older children and adults (average, ⬎ 1.5 per patient; Appendix Figs A1A and A1B, online only). Deletions of IKZF1 and BTG1 were strongly associated with IGH@-CRLF2, whereas deletions of IKZF1 were not identified in patients with ID4 involvement (data not shown). Children with or without an IKZF1, CDKN2A/B, or PAX5 deletion did not differ with respect to sex, age, WBC count, or NCI risk group (Appendix Table A7, online only). Demographic and Clinical Characterization of IGH@-Positive Patients There were no significant associations between patients with and without an IGH@ translocation with respect to sex and WBC count. JOURNAL OF CLINICAL ONCOLOGY
Information downloaded from jco.ascopubs.org and provided by at University of Brighton on July 10, 2014 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 194.81.203.94
IGH@ Translocations in ALL
A
(n = 4) (n = 5)
BCL2
(n = 4)
CEBPA
(n = 5) (n = 2) (n = 1)
CEBPB CEBPD CEBPE CEBPG CRLF2
(n = 35)
(n = 84)
EPOR ID4 IGF2BP1 IGK (n = 1)
(n = 11)
Mir-125-b TAL1
(n = 3)
B
(n = 1)
(n = 1) (n = 1)
TCRA/D
(n = 1)
Unknown
100
Positive Nuclei (%)
* 80 60 40 20
* *
0 BCL2
CEBPA CEBPB CEBPD CEBPE CRLF2
ID4
Unknown
Recurrent Partner Gene
C
12 Unknown
Other
ID4
CRLF2
CEBP
Frequency
10 8 6 4 2 0
32
䡠
NUMBER
14
䡠
MAY
10
2014
JOURNAL OF CLINICAL ONCOLOGY
O R I G I N A L
R E P O R T
IGH@ Translocations Are Prevalent in Teenagers and Young Adults With Acute Lymphoblastic Leukemia and Are Associated With a Poor Outcome Lisa J. Russell, Amir Enshaei, Lisa Jones, Amy Erhorn, Dino Masic, Helen Bentley, Karl S. Laczko, Adele K. Fielding, Anthony H. Goldstone, Nicholas Goulden, Christopher D. Mitchell, Rachel Wade, Ajay Vora, Anthony V. Moorman, and Christine J. Harrison Lisa J. Russell, Amir Enshaei, Lisa Jones, Amy Erhorn, Dino Masic, Helen Bentley, Anthony V. Moorman, and Christine J. Harrison, Leukaemia Research Cytogenetics Group, Northern Institute for Cancer Research, Newcastle University, Newcastle-uponTyne; Karl S. Laczko, Leica Microsystems, Gateshead; Adele K. Fielding, Royal Free and University College London Medical School; Anthony H. Goldstone, University College London Hospital; Nicholas Goulden, Great Ormond St Hospital, London; Christopher D. Mitchell, John Radcliffe Hospital; Rachel Wade, Clinical Trial Service Unit, University of Oxford, Oxford; and Ajay Vora, Sheffield Children’s Hospital, Sheffield, United Kingdom. Published online ahead of print at www.jco.org on April 7, 2014. Supported by the Kay Kendall Leukemia Fund and Leukemia & Lymphoma Research (formerly Leukemia Research). A.V.M. and C.J.H. share senior authorship. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article. Corresponding author: Lisa J. Russell, PhD, Leukaemia Research Cytogenetics Group, Northern Institute for Cancer Research, Level 5, Sir James Spence Institute, Royal Victoria Infirmary, Newcastle-upon-Tyne, NE1 4LP, United Kingdom; e-mail: lisa.russell@ newcastle.ac.uk. © 2014 by American Society of Clinical Oncology 0732-183X/14/3214w-1453w/$20.00 DOI: 10.1200/JCO.2013.51.3242
A
B
S
T
R
A
C
T
Purpose To determine the prevalence and prognostic association of immunoglobulin heavy chain (IGH@) translocations in acute lymphoblastic leukemia (ALL). Patients and Methods The cohort comprised 3,269 patients treated on either the UKALL2003 trial for children and adolescents (1 to 24 years old) or the UKALLXII trial for adolescents and adults (15 to 59 years old). High-throughput fluorescent in situ hybridization was used to detect IGH@ translocations. Results We identified IGH@ translocations in 5% of patients with ALL (159 of 3,269 patients), in patients with both B-cell (148 of 2,863 patients) and T-cell (11 of 408 patients) disease. Multiple partner genes were identified including CRLF2 (n ⫽ 35), five members of the CEPB gene family (n ⫽ 17), and ID4 (n ⫽ 11). The level of the IGH@-positive clone varied and indicated that some IGH@ translocations were primary events, whereas others were secondary aberrations often associated with other established aberrations. The age profile of patients with IGH@ translocations was distinctive, with a median age of 16 years and peak incidence of 11% among 20- to 24-year-old patients. Among patients with B-cell precursor ALL who were Philadelphia chromosome negative, those with an IGH@ translocation had an inferior overall survival compared with other patients (UKALL2003: hazard ratio, 2.37; 95% CI, 1.34 to 4.18; P ⫽ .003; UKALLXII: hazard ratio, 1.73; 95% CI, 1.22 to 2.47; P ⫽ .002). However, this adverse effect was not independent of age or minimal residual disease status and did not seem to be driven by an increased risk of relapse. Conclusion IGH@ translocations define a genetic feature that is frequent among adolescents and young adults with ALL. Although associated with an adverse outcome in adults, it is not an independent prognostic factor in children and adolescents. J Clin Oncol 32:1453-1462. © 2014 by American Society of Clinical Oncology
INTRODUCTION
Genetic aberrations are important indicators of prognosis in acute lymphoblastic leukemia (ALL) and are widely used in risk stratification.1 Translocations involving the immunoglobulin heavy chain locus (IGH@) are hallmarks of mature B-cell malignancies, where they drive pathogenesis (reviewed in Molyneux et al2 and Freedman3). IGH@ translocations have been described in B-cell precursor ALL (BCP-ALL), where they target different genes with the same consequence; the partner gene is overexpressed as a result of its close proximity to the IGH@ enhancer. Previously reported recurrent BCP-ALL translocation partner genes include five members of
the CCAAT/enhancer binding protein (CEBP) family of transcription factors, showing opposing functions for deregulation in myeloid and lymphoid leukemogenesis4,5; the inhibitory transcription factor, ID4, in the translocation, t(6;14)(p22;q32), defining a subgroup characterized by deletion of CDKN2A/B and PAX56; the cytokine receptor for erythropoietin (EPOR) at 19p13; type I cytokine receptor, CRLF2, at Xp22/Yp117,8; and other sporadic partner genes.9-12 We recently reported the incidence and clinical relevance of IGH@ translocations in Philadelphia chromosome–negative adult BCPALL.13 However, the incidence, spectrum, and prognostic relevance of IGH@ translocations in © 2014 by American Society of Clinical Oncology
Information downloaded from jco.ascopubs.org and provided by at University of Brighton on July 10, 2014 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 194.81.203.94
1453
Russell et al
ALL across all age groups remained unknown. In this study, we developed a high-throughput fluorescent in situ hybridization (FISH) screening approach to ascertain the incidence of these translocations among 3,269 patients and examined their prognostic relevance.
PATIENTS AND METHODS Patients Between October 2003 and June 2011, 3,182 children and adolescents, age 1 to 24 years at diagnosis, were treated on the UKALL2003 trial. Between January 1993 and November 2008, 1,380 adolescents and adults, age 15 to 55 years, were treated on UKALLXII. Together, 4,562 patients were eligible for this study (Fig 1). Clinical details, including age, sex, WBC count, immunophenotype, and survival data, were collected by the Clinical Trial Service Unit, University of Oxford (Oxford, United Kingdom). Institutional review board approval was obtained at each of the collaborating centers. Informed consent was obtained in accordance with the Declaration of Helsinki. In UKALL2003, minimal residual disease (MRD) was evaluated by realtime quantitative polymerase chain reaction analysis of immunoglobulin and T-cell receptor gene rearrangements as defined by the European MRD Study
Group.14 Patients were classified by MRD at day 28 (end of induction) as high (⬎ 0.01%) or low (⬍ 0.01%) risk. Cytogenetics and FISH Fixed cells were available from 3,302 patients, including 2,595 children (2,279 with BCP-ALL, 54 with Down syndrome and ALL [DS-ALL], and 316 with T-cell ALL [T-ALL]) and 707 adults (615 with BCP-ALL and 92 with T-ALL; Fig 1). Patients were screened for IGH@ translocations using a high-throughput automated capture and scoring FISH approach or FISH by eye (Appendix, online only). The involvement of IGH@ was determined using a variety of IGH@ FISH probes (Appendix Table A1, online only). Images were captured and scored using Cytovision SPOT counting software (Leica Microsystems, Gateshead, United Kingdom). Because IGH@-CRLF2 is cryptic, FISH for rearrangements involving CRLF2 was performed on patients with an IGH@-positive FISH result.8 Patients with known, cytogenetically visible translocations were tested with appropriate FISH probes, as previously reported.4-8,12 Multiplex Ligation-Dependent Probe Amplification DNA extracted directly from the bone marrow aspirate or fixed cells used for FISH was available from 58 UKALL2003 and 30 UKALLXII IGH@-positive patients with BCP-ALL. Copy number alterations were investigated by multiplex ligation-dependent probe amplification (MLPA) using the SALSA MLPA kit P335-A1 (MRC Holland, Amsterdam, the Netherlands).15 The kit includes probes to IKZF1, CDKN2A/B, PAX5, EBF1, ETV6, BTG1, RB1, and CSF2RA/IL3RA/CRLF2.
Total patients (N = 4,562) Excluded FISH not tested
(n =1,260)
Available for incidence analysis (n = 3,302)
UKALL2003 trial IGH@ negative IGH@ positive
(n = 2,595) (n = 2,469) (n = 103)
UKALLXII trial IGH@ negative IGH@ positive
(n = 707) (n = 641) (n = 56) Excluded (n = 2,179) IGH@- and IGH status not known, MLPA not tested
Excluded (n = 2,179) IGH@- and IGH status not known, MLPA not tested MLPA analysis (n = 58)
MLPA analysis (n = 30)
Excluded (n = 258) Lost to follow-up, t(9;22) and T cell
Excluded (n = 406) Lost to follow-up, t(9;22) and T cell Survival analysis (n = 2,194)
Survival analysis (n = 449)
Fig 1. Analysis profile for investigation of the incidence of IGH@ translocations, survival analysis, and prevalence of specific gene deletions. A total of 4,562 patients were entered onto two treatment trials, the UKALL2003 and UKALLXII trials. Fixed material was available from 3,302 patients, including 2,595 childhood patients from UKALL2003 and 707 adults from UKALLXII; the remaining 1,260 patients without fixed material were excluded. Before multiplex ligation-dependent probe amplification (MLPA) or survival analysis, patients who did not have an IGH@ translocation, had T-cell disease, or had Philadelphia chromosome–positive acute lymphoblastic leukemia were excluded from further analysis. From a total of 103 childhood and 56 adult IGH@-positive patients, DNA was available from 58 and 30 of them, respectively, for MLPA analysis. Survival data were available from 2,194 children and 449 adults who had a known IGH@ result, both positive and negative. Patients with no follow-up data were removed before analysis. FISH, fluorescent in situ hybridization. 1454
© 2014 by American Society of Clinical Oncology
JOURNAL OF CLINICAL ONCOLOGY
Information downloaded from jco.ascopubs.org and provided by at University of Brighton on July 10, 2014 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 194.81.203.94
IGH@ Translocations in ALL
Table 1. Demographics and Clinical Characteristics of UKALL2003 Patients With and Without IGH@ Translocations Demographic or Clinical Characteristic Total Sex Female Male Age, years 1-4 5-9 10-14 15-19 20-24 Median WBC count, ⫻ 109/L ⬍ 50 ⬎ 50 Median NCI risk groupⴱ Standard risk High risk Immunophenotype B cell T cell DS-ALL No Yes MRD risk group† High Low Ph status Negative Positive
Total Patients
IGH@-Negative Patients
Patients With Failed Tests
IGH@-Positive Patients
No.
%
No.
% of Total
No.
% of Total
2,595
100
2,469
95
103
4
P
No.
% of Total
23
1
8 15
0 0
13 7 1 1 1
1 1 0 0 2
.42 1,112 1,483
100 100
1,064 1,405
96 95
40 63
4 4
1,246 649 430 221 49
100 100 100 100 100
1,212 622 402 194 39
97 96 93 88 80
21 20 27 26 9
2 3 6 12 18
⬍ .001
5
5
12
5 .91
1,991 604
100 100
1,892 577
13
95 96
80 23
14
4 4
19 4
12
0 0 7
⬍ .001 1,486 1,109
100 100
1,437 1,032
97 93
33 70
2 6
2,279 316
100 100
2,162 307
95 97
95 8
4 3
2,538 57
100 100
2,423 46
95 81
94 9
4 16
911 849
100 100
880 790
97 93
21 53
2 6
2,541 54
100 100
2,417 52
95 96
101 2
4 4
16 7
1 1
22 1
0 0
21 2
1 3
10 6
1 1
23 0
1 0
.22
.02
⬍ .001
1.0
Abbreviations: DS-ALL, Down syndrome and acute lymphoblastic leukemia; MRD, minimal residual disease; NCI, National Cancer Institute; Ph, Philadelphia translocation. ⴱ NCI risk stratifies risk according to age and WBC count. Standard risk indicates WBC count less than 50,000/mL and age 1 to younger than 10 years. High risk indicates WBC count ⱖ 50,000/mL and age ⱖ 10 years. †MRD was evaluated by real-time quantitative polymerase chain reaction analysis of immunoglobulin and T-cell receptor gene rearrangements as defined by the European MRD Study Group. Patients were stratified by MRD at the end of induction into high-risk (⬎ 0.01%) and low-risk (⬍ 0.01%) groups. MRD risk group could not be assigned for 728 patients. Data on MRD are not available for 837 patients.
Survival Survival analysis was restricted to patients with BCP-ALL, and separate analyses were performed for UKALL2003 and UKALLXII. Philadelphia chromosome–positive patients were excluded from analysis, because many received imatinib. Therefore, 2,194 children and young adults (UKALL2003) and 449 adults (UKALLXII) were analyzed (Fig 1). Survival analysis considered the following three end points: event-free survival, defined as time to relapse or death or second neoplasm, with censoring at last contact; relapse rate, defined as time from complete remission (CR) to relapse, with censoring at death in remission or last contact; and overall survival, defined as time to death, with censoring at last contact. Patients who did not achieve a CR were included in the analysis until an event or censor. Survival rates were calculated and compared using the Kaplan-Meier method, log-rank test, and Cox regression models (univariate and multivariable analyses). The Fisher’s exact test was used to examine the distribution of categorical variables. Because of the investigative nature of this analysis, we did not apply stringent multiple comparisons adjustment. However, all tests were conducted at the 1% significance level. All analyses were performed using Intercooled Stata 11.0 for Windows (Stata Corporation, College Station, TX). www.jco.org
RESULTS
Genetic Characterization of IGH@ Translocations Incidence of IGH@ translocations and the partner genes. IGH@ FISH was successful for 3,269 patients; 33 patients (UKALL2003, n ⫽ 23; UKALLXII, n ⫽ 10) had two failed independent tests (Tables 1 and 2). Patients who were tested did not differ from those not tested with respect to sex, age, WBC count, and National Cancer Institute (NCI) risk group (Appendix Tables A2 and A3, online only). However, in the adult cohort, fewer patients with T-ALL were tested (P ⬍ .002; Appendix Table A3). IGH@ translocations were observed in 5% of patients (159 of 3,269 patients; Tables 1 and 2), including 148 (5%) of 2,863 patients with BCP-ALL and 11 (3%) of 407 patients with T-ALL. CRLF2 was the most prevalent IGH@ translocation partner gene, identified in 35 (22%) of 159 IGH@-positive patients but not in patients with T-ALL (Appendix Table A4, online only; Fig 2A). The © 2014 by American Society of Clinical Oncology
Information downloaded from jco.ascopubs.org and provided by at University of Brighton on July 10, 2014 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 194.81.203.94
1455
Russell et al
Table 2. Demographics and Clinical Characteristics for UKALLXII Adult Patients With and Without IGH@ Translocations Demographic or Clinical Characteristic Total Sex Female Male WBC count, ⫻ 109/L ⬍ 50 ⬎ 50 Median Age, years 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60⫹ Median Immunophenotype B cell T cell Ph status Negative Positive
IGH@-Negative Patients
Total Patients
Patients With Failed Tests
IGH@-Positive Patients
No.
%
No.
% of Total
No.
% of Total
707
100
641
91
56
8
P
No.
% of Total
10
1
4 6
1 1
8 2
1 1
.57 288 419
100 100
259 382
90 91
25 31
9 7
518 180
100 100
471 162
91 90
39 16
8 9
.63
14
14
15
7 .98
119 95 79 60 60 74 69 54 21 10
100 100 100 100 100 100 100 100 100 100
119 95 79 60 60 74 69 54 21 10
33
89 91 93 90 91 94 87 92 91 83
12 8 5 6 5 4 9 3 2 2
33
9 8 6 9 8 5 11 5 9 17
2 1 1 1 1 1 1 2 0 0
33
2 1 1 1 1 1 1 2 0 0 33
.1 615 92
100 100
553 88
90 96
53 3
9 3
554 153
100 100
494 147
89 96
52 4
9 3
9 1
1 1
8 2
1 1
⬍ .001
Abbreviation: Ph, Philadelphia translocation.
CEBP gene family and ID4 were involved in 17 (11%) of 159 and 11 (7%) of 159 IGH@-positive BCP-ALLs, respectively (Fig 2A), and were also not identified in T-ALL. BCL2 was involved in four patients, and other partner genes were rare (ⱕ two patients each) (Appendix Table A4, Fig 2A). In T-ALL, one novel partner was identified (TAL1). The involvement of TRA/D@ was observed in both B- and T-cell disease. The percentage of IGH@-positive nuclei ranged from 5% to 100% and varied according to the partner gene (Fig 2B). For some partners (BCL2, CEBPA, CEBPB, and CRLF2), rearrangements were present in the majority of nuclei (median, 79%), whereas the proportion was lower for other genes (ID4: median, 65%; CEPBE: median, 30%; and unknown partner genes: median, 30%). Cytogenetic characterization of IGH@-positive patients. Among IGH@-positive patients (n ⫽ 159), 37% had a normal karyotype (n ⫽ 43) or a failed cytogenetics test (n ⫽ 16). The IGH@ translocation was the sole abnormality in 11 patients, whereas additional cytogenetic changes were present in the majority of patients (89 of 100 patients; Fig 3; Appendix Table A4). IGH@ translocations were observed with an established cytogenetic aberration in 28 (19%) of 148 patients with BCP-ALL, including BCR-ABL1 (n ⫽ 6), intrachromosomal amplification of chromosome 21 (iAMP21; n ⫽ 3), MLL rearrangements (n ⫽ 2), high hyperdiploidy (n ⫽ 9), and ETV6-RUNX1 (n ⫽ 8). The partner gene was known in only six of these patients (Appendix Table A4). In T-ALL, single IGH@-positive patients had rearrangements of STIL-TAL1, 1456
© 2014 by American Society of Clinical Oncology
MLL, TAL2, MLLT10-PICALM, LMO2, HOX11, BCL11B, TLX3, TRB@, and TRA/D@ (Appendix Table A4). Copy number abnormalities. We assessed the copy number status of eight genes (IKZF1, CDKN2A/B, PAX5, EBF1, ETV6, BTG1, and RB1) and the presence of P2RY8-CRLF2 fusion in 88 IGH@-positive patients with BCP-ALL using MLPA (UKALL2003, n ⫽ 58; UKALLXII, n ⫽ 30). There were no differences in sex, age, and WBC count between IGH@-positive patients tested and not tested by MLPA (Appendix Table A5, online only). Deletions were identified in all eight genes, with CDKN2A/B and IKZF1 deletions being the most prevalent (approximately 40%; Appendix Table A6, online only). Deletions were rarer in children age 1 to 9 years (average, 0.84 per patient) compared with older children and adults (average, ⬎ 1.5 per patient; Appendix Figs A1A and A1B, online only). Deletions of IKZF1 and BTG1 were strongly associated with IGH@-CRLF2, whereas deletions of IKZF1 were not identified in patients with ID4 involvement (data not shown). Children with or without an IKZF1, CDKN2A/B, or PAX5 deletion did not differ with respect to sex, age, WBC count, or NCI risk group (Appendix Table A7, online only). Demographic and Clinical Characterization of IGH@-Positive Patients There were no significant associations between patients with and without an IGH@ translocation with respect to sex and WBC count. JOURNAL OF CLINICAL ONCOLOGY
Information downloaded from jco.ascopubs.org and provided by at University of Brighton on July 10, 2014 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 194.81.203.94
IGH@ Translocations in ALL
A
(n = 4) (n = 5)
BCL2
(n = 4)
CEBPA
(n = 5) (n = 2) (n = 1)
CEBPB CEBPD CEBPE CEBPG CRLF2
(n = 35)
(n = 84)
EPOR ID4 IGF2BP1 IGK (n = 1)
(n = 11)
Mir-125-b TAL1
(n = 3)
B
(n = 1)
(n = 1) (n = 1)
TCRA/D
(n = 1)
Unknown
100
Positive Nuclei (%)
* 80 60 40 20
* *
0 BCL2
CEBPA CEBPB CEBPD CEBPE CRLF2
ID4
Unknown
Recurrent Partner Gene
C
12 Unknown
Other
ID4
CRLF2
CEBP
Frequency
10 8 6 4 2 0
