SUPPLEMENTAL DATAfor
Prognostic Value of Rare IKZF1 Deletion in Childhood B-cell Precursor Acute Lymphoblastic Leukemia: An International Collaborative Study
Judith M. Boer1*, Arian van der Veer1*, Dimitris Rizopoulos2, Marta Fiocco3,4, Edwin Sonneveld3, Hester A. de Groot-Kruseman3, Roland P. Kuiper5,Peter Hoogerbrugge3,6, Martin Horstmann7, Marketa Zaliova8,9, Chiara Palmi10, Jan Trka9, Eva Fronkova9, Mariana Emerenciano11, Maria do Socorro Pombo-de-Oliveira11, Wojciech Mlynarski12, Tomasz Szczepanski13, Karin Nebral14, Andishe Attarbaschi15, Nicola Venn16, Rosemary Sutton16, Claire J. Schwab17, Amir Enshaei17,Ajay Vora18, Martin Stanulla19, Martin Schrappe8, Gianni Cazzaniga10, Valentino Conter10, Martin Zimmermann19, Anthony V. Moorman17, Rob Pieters1,3,6, Monique L. den Boer1,3
1Department of Pediatric Oncology, Erasmus MC – Sophia Children’s Hospital, Rotterdam, the Netherlands
2Department of Biostatistics, Erasmus MC, Rotterdam, the Netherlands
3DCOG - Dutch Childhood Oncology Group, the Hague, the Netherlands
4Department of Medical Statistics and Bioinformatics, Leiden University Medical Centre, the Netherlands
5Department of Human Genetics, Radboud University Medical Center and Radboud Institute for Molecular Life Sciences, Nijmegen, the Netherlands
6Princess Máxima Center for Pediatric Oncology, Utrecht, the Netherlands
7COALL study group, Research Centre and Clinic for Pediatric Oncology, University Medical Centre Eppendorf, Hamburg, Germany
8BFM-G - Berlin-Frankfurt-Münster-Germany study group, University of Schleswig-Holstein, Kiel, Germany
9CLIP – Childhood Leukaemia Investigation Prague, Second Faculty of Medicine, Charles University and University Hospital Motol, Prague, Czech Republic
10AIEOP - Associazione Italiana Ematologia Oncologia Pediatrica, University of Milano‐Bicocca, Monza, Italy
11Pediatric Haematology and Oncology Program, Research Centre, Instituto Nacional de Câncer, Rio de Janeiro, Brazil
12PPLLSG - Polish Pediatric Leukaemia Lymphoma Study Group, Department of Pediatrics, Oncology, Haematology and Diabetology, Medical University of Lodz, Lodz, Poland
13PPLLSG - Polish Pediatric Leukaemia Lymphoma Study Group, Department of Pediatric Hematology and Oncology, Medical University of Silesia, Zabrze, Poland
14Children's Cancer Research Institute, St. Anna Kinderkrebsforschung, Vienna, Austria
15BFM-A - Berlin-Frankfurt-Münster-Austria study group, Department of Pediatric Hematology and Oncology, St. Anna Children's Hospital, Department of Pediatrics and Adolescent Medicine, Medical University of Vienna, Austria
16ANZCHOG – Australian and New Zealand Children’s Oncology Group, Children’s Cancer Institute Australia, Lowy Cancer Research Centre, UNSW, Sydney, Australia
17Leukaemia Research Cytogenetics Group, Northern Institute for Cancer Research, Newcastle University, Newcastle-upon-Tyne, UK
18Department of Haematology, Sheffield Children’s Hospital, Sheffield, UK
19Pediatric Hematology and Oncology, Hannover Medical School, Hannover, Germany
*These authors contributed equally to this work
Supplemental Methods
Inclusion criteria
All types of deletions in IKZF1, with the exception of whole gene deletions (DEL 1-8) and isoform 6 (DEL 4-7), were called rare deletions in the present study since each of these variants comprised less than 10% of IKZF1 deletions in children with BCP-ALL.1 We considered homozygous common deletions as rare variants: four biallelic DEL1-8 and one biallelic DEL4-7 were included in the study. From the MLPA data, indirect information on two different (compound) heterozygous deletions co-occurring in the same patient can be obtained. We observed two compound heterozygous deletions for rare variants (DEL4-8/DEL1 and DEL1-5/DEL1) and two compound heterozygous deletions for one common and one rare variant (DEL1-8/DEL1 and DEL1-8/DEL4-6). The nine homozygous/compound heterozygous deleted cases are indicated Figure 4 and were analyzed in the DEL-Other group.
All cases included in the study were tested for MLL rearrangements and BCR-ABL1 translocation. BCR-ABL1/MLL-rearrangement negative BCP-ALL cases that were not tested for ETV6-RUNX1, high hyperdiploidy (>50 chromosomes and/or DNA index ≥1.16), and TCF3 rearrangements were allocated to the unclassified group. Cases that were negatively tested for ETV6-RUNX1, high hyperdiploidy, TCF3 rearrangements, BCR-ABL1, and MLL rearrangements were allocated to the cytogenetic B-other group. Intrachromosomal amplification of chromosome 21 (iAMP21) was not consistently screened for in all participating centres. We, therefore, did not include iAMP21 as a separate cytogenetic subtype but included them in the B-other group. From the cases with iAMP21 data available, 2% (3/134) are among the rare IKZF1-deleted cases and 2.7% (11/402) among the IKZF1 wild-type controls. Not all centers performed microarray expression profiling necessary to identify BCR-ABL1-like cases2 among BCP-ALL cases without recurrent aberrations, therefore, this subtype was not included in the matching. The cytogenetic subtypesBCR-ABL1 and MLL-rearranged were not included in our study. Due to the low frequency of these aberrations, insufficient cases or controls could be selected. As cases and controls were matched on major cytogenetic subtype, their inclusion/exclusion will not affect the results.
IKZF1 deletions
IKZF1 deletions were identified using the SALSA P335 ALL-IKZF1 or P202 IKZF1 Multiplex Ligation-dependent Probe Amplification (MLPA) assay (MRC-Holland, Amsterdam, the Netherlands) using 125 ng of genomic DNA and following the manufacturer’s protocol. The amplified products were quantified using an ABI-3130 genetic analyzer (Applied Biosystems, Carlsbad, CA). Peak intensities—based on incorporated FAM-labeled nucleotides—were normalized to the manufacturer’s control probes and to reference DNA obtained from healthy volunteers. A ratio lower than 0.75 for the peak intensity per IKZF1 probe in patients versus the healthy control reference was considered to represent a mono-allelic deletion, and a ratio lower than 0.25 was considered to represent a bi-allelic deletion. A ratio between 0.75 and 1.3 was considered to represent a normal copy number, and a ratio higher than 1.3 was considered to represent a gain in copy number.
Study design rationale
A case-control design is the most appropriate approach if cases are rare and a multi-center/multi-national setup is required3. The study design comparing rare IKZF1 deletions (cases) vs. IKZF1 wild-type (controls) is a compromise between having sufficient samples for each rare IKZF1 variant, and being able to match for additional variablesincluding treatment protocol. Since part of the participating centers used the P202 MLPA assay, we do not have information on alterations in other B-cell development genes for all study patients and these were not taken into account.
Within this design, we collected the largest possible cohort within 10 participating international study groups consisting of in total 134 rare IKZF1 cases and 402 matched controls. Rare variant groups that were considered separately were DEL2-3, DEL2-7, DEL2-8 and DEL4-8. Of these rare variant types, sufficient cases could be included to observe at least 8 events. Remaining rare variants were grouped in the DEL-Other category. The low number of events does not allow for multivariate testing and makes a matched case-control design the best option.
Cases with rare IKZF1 deletions matched for treatment protocol, stratification arm, cytogenetic group and diagnosis date showed a significantly higher proportion of patients with white blood cell count (WBC) ≥50x109 cells/l (Chi-square p=0.005) and age ≥10 (Chi-square p=0.003). Therefore, WBC and age were included as additional matching criteria in the case-control design. Subsequently, the rare IKZF1 deletion variants and their matched controls are not expected to be representative for their entire study cohorts.
Event-free survival (EFS) was estimated using the actuarial Kaplan-Meier method. Relapse or death were counted as event. There were 3/134 rare IKZF1-deleted cases who died in complete remission (2%), and 14/402 matched controls (3%).Induction failure (non-response) was not recorded uniformly over all treatment protocols, and because of this inconsistency was left out as event in the EFS analysis. In total, our study included eight patients (4 cases, 4 controls) who were reported to have reached no complete remission at the end of induction therapy. These eight reported patients either suffered from a relapse (4) or died early during induction therapy due to leukemia (4). Because of the lack of uniformity of the definition of non-response, we counted these reported patients as event at the time of relapse or early death. A Cox's regression model was used in which the standard errors have been calculated using the jack-knife sandwich estimator to account for the matching nature of the data.4 This appropriate matched analysis was only possible in the Cox model, not in the cumulative incidence of relapse with competing risks model using the Gray test.
Case-control selection
The flowchart shows the selection of IKZF1 rare variant casesand IKZF1 wild-type controls. IKZF1 was not a stratifying factor in any of the protocols.
References for Supplemental Methods
1.van der Veer A, Waanders E, Pieters R, et al. Independent prognostic value of BCR-ABL1-like signature and IKZF1 deletion, but not high CRLF2 expression, in children with B-cell precursor ALL. Blood. 2013;122(15):2622-2629.
2.Den Boer ML, van Slegtenhorst M, De Menezes RX, et al. A subtype of childhood acute lymphoblastic leukaemia with poor treatment outcome: a genome-wide classification study. Lancet Oncol. 2009;10(2):125-134.
3.Grisso JA. Making comparisons. Lancet. 1993;342(8864):157-160.
4.Therneau T, Grambsch P. Modeling Survival Data: Extending the Cox Model. Springer-Verlag New York. 2000;Chapter 8.
Supplemental TABLES
Supplemental Table S2. Event-free survival of rare IKZF1 deletion variants in relation to risk groups, subtypes and deletion types.
A / Cases / Controls / paired EFSRisk groups / N tot / N rel / N event / N tot / N rel / N event / HR / 95% CI / p-value
standard risk / 56 / 19 / 20 / 168 / 27 / 34 / 1.9 / 1.2-3.2 / 0.01
medium risk / 33 / 10 / 10 / 99 / 16 / 17 / 1.8 / 1.3-2.5 / <0.001
high risk / 45 / 23 / 255 / 135 / 44 / 50 / 1.8 / 1.2-2.7 / 0.007
B / Cases / Controls / paired EFS
Subtypes / N tot / N rel / N event / N tot / N rel / N event / HR / 95% CI / p-value
B-other / 88 / 40 / 43 / 264 / 68 / 80 / 1.8 / 1.3-2.3 / <0.001
high hyperdiploid / 21 / 6 / 6 / 63 / 9 / 9 / 2.4 / 0.8-6.6 / 0.1
ETV6-RUNX1 / 12 / 0 / 0 / 36 / 3 / 4 / ND / ND / ND
C / Cases / Controls / paired EFS
DEL-Other types / N tot / N rel / N event / N tot / N rel / N event / HR / 95% CI / p-value
Haplo-insufficient / 15 / 6 / 8 / 45 / 11 / 11 / 2.4 / 1.1-5-5 / 0.04
Miscellaneous / 19 / 9 / 9 / 57 / 13 / 14 / 2.1 / 1.4-3.2 / <0.001
* N tot: total number of patients
** N rel: number of patients with relapse
*** N event: number of patients with event
Supplemental Table S3. Outcome of rare IKZF1 deletion variants compared with the outcome of known poor prognostic major types of IKZF1 deletions.
Variant type / EFS vs. DEL 1-8 / EFS vs. DEL 4-7 / EFS vs. DEL 1-8 + DEL 4-7N tot / N event / HR / 95% CI / p-value / HR / 95% CI / p-value / HR / 95% CI / p-value
All rare DEL / 134 / 55 / 1.1 / 0.8-1.6 / 0.7 / 0.9 / 0.6-1.4 / 0.7 / 1.0 / 0.7-1.4 / 1
DEL 2-3 / 26 / 8 / 0.8 / 0.4-1.6 / 0.5 / 0.8 / 0.4-1.6 / 0.5 / 0.8 / 0.4-1.6 / 0.5
DEL 2-7 / 32 / 12 / 1.0 / 0.5-1.8 / 1 / 1.0 / 0.5-1.8 / 1 / 1.0 / 0.5-1.8 / 1
DEL 2-8 / 15 / 10 / 2.0 / 1.0-4.0 / 0.04 / 2.0 / 1.0-4.0 / 0.04 / 2.0 / 1.0-4.0 / 0.04
DEL 4-8 / 27 / 8 / 0.8 / 0.4-1.6 / 0.5 / 0.8 / 0.4-1.6 / 0.5 / 0.8 / 0.4-1.6 / 0.5
DEL-Other / 34 / 17 / 1.3 / 0.7-2.2 / 0.4 / 1.3 / 0.7-2.2 / 0.4 / 1.3 / 0.7-2.2 / 0.4
DEL 1-8 / 163 / 60 / 1.0
DEL 4-7 / 146 / 55 / 1.0
DEL 4-7 + DEL 1-8 / 309 / 115 / 1.0
Supplemental figures
Supplemental Figure S1. Localization of MLPA probes.IKZF1 gene is depicted in bars. Transparant bars represent non-coding sequences. Gray bars indicate amino acid coding sequences. Zinc fingers are depicted in black. The DNA-binding, activation and dimerization regions are assigned by double arrows. Normal arrows indicate the localization of SALSA MLPA P202 and SALSA MLPA P335 probes respectively.
Supplemental Figure S2. Event-free survival for all rare IKZF1-deleted cases compared with their wild-type controls per treatment protocol stratification arm. IKZF1 rare deletion case were compared to matched IKZF1 wild-type controls for (A) standard risk (SR), (B) medium risk (MR), (C) and high risk (HR) treated patients. The total number of patients (n), and the number of events (e; relapse or death) are indicated for each curve.For statistical tests see Supplemental Table S2A.
Supplemental Figure S3. Event-free survival for all rare IKZF1-deleted cases compared with their matched wild-type controls per cytogenetic group.(A) B-other, i.e. negative for BCR-ABL1, MLL-rearrangements, TCF3 rearrangements, ETV6-RUNX1 and high hyperdiploidy. (B) High hyperdiploidy (51-65 chromosomes). (C) ETV6-RUNX1; P-values for the ETV6-RUNX1-positive groups were not determined (ND), since no events occurred in the DEL group. The total number of patients (n), and the number of events (e; relapse or death) are indicated for each curve. For statistical tests see Supplemental Table S2B.
Supplemental Figure S4. EFS curves for IKZF1 DEL-Other group compared with matched wild-type controls. (A) Haplo-insufficient DEL-Other cases, including the translational start site in exon 2 in the deletion, and (B) miscellaneous DEL-Other cases. Total patients (n), and events (e; relapse or death) are indicated for each curve. For statistical tests see Supplemental Table S2C.
1