Automated identification of cell populations in flow cytometry data with transformers.

Acute Lymphoblastic Leukemia (ALL) is the most frequent hematologic malignancy in children and adolescents. A strong prognostic factor in ALL is given by the Minimal Residual Disease (MRD), which is a measure for the number of leukemic cells persistent in a patient. Manual MRD assessment from Multiparameter Flow Cytometry (FCM) data after treatment is time-consuming and subjective. In this work, we present an automated method to compute the MRD value directly from FCM data. We present a novel neural network approach based on the transformer architecture that learns to directly identify blast cells in a sample. We train our method in a supervised manner and evaluate it on publicly available ALL FCM data from three different clinical centers. Our method reaches a median F1 score of ≈0.94 when evaluated on 519 B-ALL samples and shows better results than existing methods on 4 different datasets.

An Extensive Quality Control and Quality Assurance (QC/QA) Program Significantly Improves Inter-Laboratory Concordance Rates of Flow-Cytometric Minimal Residual Disease Assessment in Acute Lymphoblastic Leukemia: An I-BFM-FLOW-Network Report.

Monitoring of minimal residual disease (MRD) by flow cytometry (FCM) is a powerful prognostic tool for predicting outcomes in acute lymphoblastic leukemia (ALL). To apply FCM-MRD in large, collaborative trials, dedicated laboratory staff must be educated to concordantly high levels of expertise and their performance quality should be continuously monitored. We sought to install a unique and comprehensive training and quality control (QC) program involving a large number of reference laboratories within the international Berlin-Frankfurt-Münster (I-BFM) consortium, in order to complement the standardization of the methodology with an educational component and persistent quality control measures. Our QC and quality assurance (QA) program is based on four major cornerstones: (i) a twinning maturation program, (ii) obligatory participation in external QA programs (spiked sample send around, United Kingdom National External Quality Assessment Service (UK NEQAS)), (iii) regular participation in list-mode-data (LMD) file ring trials (FCM data file send arounds), and (iv) surveys of independent data derived from trial results. We demonstrate that the training of laboratories using experienced twinning partners, along with continuous educational feedback significantly improves the performance of laboratories in detecting and quantifying MRD in pediatric ALL patients. Overall, our extensive education and quality control program improved inter-laboratory concordance rates of FCM-MRD assessments and ultimately led to a very high conformity of risk estimates in independent patient cohorts.

Automated Flow Cytometric MRD Assessment in Childhood Acute B- Lymphoblastic Leukemia Using Supervised Machine Learning.

Minimal residual disease (MRD) as measured by multiparameter flow cytometry (FCM) is an independent and strong prognostic factor in B-cell acute lymphoblastic leukemia (B-ALL). However, reliable flow cytometric detection of MRD strongly depends on operator skills and expert knowledge. Hence, an objective, automated tool for reliable FCM-MRD quantification, able to overcome the technical diversity and analytical subjectivity, would be most helpful. We developed a supervised machine learning approach using a combination of multiple Gaussian Mixture Models (GMM) as a parametric density model. The approach was used for finding the weights of a linear combination of multiple GMMs to represent new, "unseen" samples by an interpolation of stored samples. The experimental data set contained FCM-MRD data of 337 bone marrow samples collected at day 15 of induction therapy in three different laboratories from pediatric patients with B-ALL for which accurate, expert-set gates existed. We compared MRD quantification by our proposed GMM approach to operator assessments, its performance on data from different laboratories, as well as to other state-of-the-art automated read-out methods. Our proposed GMM-combination approach proved superior over support vector machines, deep neural networks, and a single GMM approach in terms of precision and average F 1 -scores. A high correlation of expert operator-based and automated MRD assessment was achieved with reliable automated MRD quantification (F 1 -scores >0.5 in more than 95% of samples) in the clinically relevant range. Although best performance was found, if test and training samples were from the same system (i.e., flow cytometer and staining panel; lowest median F 1 -score 0.92), cross-system performance remained high with a median F 1 -score above 0.85 in all settings. In conclusion, our proposed automated approach could potentially be used to assess FCM-MRD in B-ALL in an objective and standardized manner across different laboratories. © 2019 International Society for Advancement of Cytometry.

Flow-cytometric minimal residual disease monitoring in blood predicts relapse risk in pediatric B-cell precursor acute lymphoblastic leukemia in trial AIEOP-BFM-ALL 2000.

BACKGROUND: Flow-cytometric monitoring of minimal residual disease (MRD) in bone marrow (BM) during induction of pediatric patients with acute lymphoblastic leukemia (ALL) is widely used for outcome prognostication and treatment stratification. Utilizing peripheral blood (PB) instead of BM might be favorable, but data on its usefulness are scarce. PROCEDURE: We investigated 1303 PB samples (days 0, 8, 15, 33, and 52) and 285 BMs (day 15) from 288 pediatric ALL patients treated in trial AIEOP-BFM ALL 2000. MRD was assessed by four-color flow cytometry and evaluated as relative, absolute, and kinetic result. RESULTS: In B-ALL only, PB measures from early time points correlated with relapse incidence (CIR). Best separation occurred at threshold <1 blast/µL at day 8 (5-year CIR 0.02 ± 0.02 vs 0.12 ± 0.03; P = 0.044). Patients with highest relapse risk were not distinguishable, but PB-MRD at days 33 and 52 correlated with prednisone response and postinduction BM-MRD by PCR (P < 0.001). Kinetic assessment did not convey any advantage. In multivariate analysis including day 15 BM-MRD, PB-MRD measures lost statistical power. CONCLUSIONS: In summary, PB-MRD in pediatric B-ALL correlates with outcome and risk parameters, but its prognostic significance is not strong enough to substitute for BM assessment in AIEOP-BFM trials. It might, however, be valuable in treatment environments not using multifaceted risk stratification with other MRD measures.

Expression of CD56 defines a distinct subgroup in childhood T-ALL with inferior outcome. Results of the ALL-BFM 2000 trial.

This study reports the prognostic impact of the expression of the natural killer cell marker CD56 in a large series of risk-adapted paediatric patients with T cell acute lymphoblastic leukaemia (T-ALL; n = 493) treated within the ALL-Berlin-Frankfurt-Münster (BFM) 2000 protocol. The immunophenotype was analysed centrally at diagnosis using flow cytometry and correlated with clinical parameters and outcome. CD56 expression was detected in 7·1% and early T-cell precursor (ETP) phenotype in 6·7% of all T-ALL patients. The percentage of ETP in the CD56+ T-ALL cohort was 4-fold higher than in the whole cohort. CD56+ T-ALL frequently expressed the progenitor marker CD34 and myeloid antigens CD13 and CD33. The 5-year event-free survival (EFS) rates for the European Group for the Immunological classification of Leukaemias/World Health Organization subgroups and the ETP phenotype were not statistically different. By contrast, patients with CD56 expression had a significantly reduced EFS (60 ± 8%) and overall survival (60 ± 8%) at 5 years, with a hazard ratio of 2·46 (P = 0·002) and 2·99 (P < 0·001), respectively. Moreover, CD56 expression in combination with the minimal residual disease (MRD)-based high risk assignment defined a population with a 'very-high' risk probability of relapse in the ALL-BFM 2000 trial. The CD56 marker has the potential to augment MRD-based risk stratification and may serve as a molecular target for antibody-based treatment strategies in childhood T-ALL.

International cooperative study identifies treatment strategy in childhood ambiguous lineage leukemia.

Despite attempts to improve the definitions of ambiguous lineage leukemia (ALAL) during the last 2 decades, general therapy recommendations are missing. Herein, we report a large cohort of children with ALAL and propose a treatment strategy. A retrospective multinational study (International Berlin-Frankfurt-Münster Study of Leukemias of Ambiguous Lineage [iBFM-AMBI2012]) of 233 cases of pediatric ALAL patients is presented. Survival statistics were used to compare the prognosis of subsets and types of treatment. Five-year event-free survival (EFS) of patients with acute lymphoblastic leukemia (ALL)-type primary therapy (80% ± 4%) was superior to that of children who received acute myeloid leukemia (AML)-type or combined-type treatment (36% ± 7.2% and 50% ± 12%, respectively). When ALL- or AML-specific gene fusions were excluded, 5-year EFS of CD19+ leukemia was 83% ± 5.3% on ALL-type primary treatment compared with 0% ± 0% and 28% ± 14% on AML-type and combined-type primary treatment, respectively. Superiority of ALL-type treatment was documented in single-population mixed phenotype ALAL (using World Health Organization and/or European Group for Immunophenotyping of Leukemia definitions) and bilineal ALAL. Treatment with ALL-type protocols is recommended for the majority of pediatric patients with ALAL, including cases with CD19+ ALAL. AML-type treatment is preferred in a minority of ALAL cases with CD19- and no other lymphoid features. No overall benefit of transplantation was documented, and it could be introduced in some patients with a poor response to treatment. As no clear indicator was found for a change in treatment type, this is to be considered only in cases with ≥5% blasts after remission induction. The results provide a basis for a prospective trial.

Loss-of-function uORF mutations in human malignancies.

Ribosome profiling revealed widespread translational activity at upstream open reading frames (uORFs) and validated uORF-mediated translational control as a commonly repressive mechanism of gene expression. Translational activation of proto-oncogenes through loss-of-uORF mutations has been demonstrated, yet a systematic search for cancer-associated genetic alterations in uORFs is lacking. Here, we applied a PCR-based, multiplex identifier-tagged deep sequencing approach to screen 404 uORF translation initiation sites of 83 human tyrosine kinases and 49 other proto-oncogenes in 308 human malignancies. We identified loss-of-function uORF mutations in EPHB1 in two samples derived from breast and colon cancer, and in MAP2K6 in a sample of colon adenocarcinoma. Both mutations were associated with enhanced translation, suggesting that loss-of-uORF-mediated translational induction of the downstream main protein coding sequence may have contributed to carcinogenesis. Computational analysis of whole exome sequencing datasets of 464 colon adenocarcinomas subsequently revealed another 53 non-recurrent somatic mutations functionally deleting 22 uORF initiation and 31 uORF termination codons, respectively. These data provide evidence for somatic mutations affecting uORF initiation and termination codons in human cancer. The insufficient coverage of uORF regions in current whole exome sequencing datasets demands for future genome-wide analyses to ultimately define the contribution of uORF-mediated translational deregulation in oncogenesis.

AIEOP-BFM consensus guidelines 2016 for flow cytometric immunophenotyping of Pediatric acute lymphoblastic leukemia.

Immunophenotyping by flow cytometry (FCM) is a worldwide mainstay in leukemia diagnostics. For concordant multicentric application, however, a gap exists between available classification systems, technologic standardization, and clinical needs. The AIEOP-BFM consortium induced an extensive standardization and validation effort between its nine national reference laboratories collaborating in immunophenotyping of pediatric acute lymphoblastic leukemia (ALL). We elaborated common guidelines which take advantage of the possibilities of multi-color FCM: marker panel requirements, immunological blast gating, in-sample controls, tri-partite antigen expression rating (negative vs. weak or strong positive) with capturing of blast cell heterogeneities and subclone formation, refined ALL subclassification, and a dominant lineage assignment algorithm able to distinguish "simple" from bilineal/"complex" mixed phenotype acute leukemia (MPAL) cases, which is essential for choice of treatment. These guidelines are a first step toward necessary inter-laboratory standardization of pediatric leukemia immunophenotyping for a concordant multicentric application. © 2017 International Clinical Cytometry Society.

Results and conclusions of the European Intergroup EURO-LB02 trial in children and adolescents with lymphoblastic lymphoma.

In the European Intergroup EURO-LB02 trial, children and adolescents with lymphoblastic lymphoma underwent the non-Hodgkin lymphoma Berlin-Frankfurt-Münster protocol without prophylactic cranial radiotherapy. The primary aims of this trial were to test whether replacing prednisone with dexamethasone during induction increases event-free survival in the subgroups with T-cell lymphoblastic lymphoma and whether therapy duration could be reduced from 24 to 18 months (factorial design, randomizations). These questions could not be answered due to premature closure of the trial. Here we report on the secondary aims of the trial: whether the results of the NHL-BFM90 study could be reproduced and evaluation of disease features and prognostic factors. Three hundred and nineteen patients (66 with precursor B-cell lymphoblastic lymphoma, 233 with T-cell lymphoblastic lymphoma, 12 with mixed phenotype, 8 not classifiable) were enrolled. In induction, 215 patients received prednisone and 104 patients received dexamethasone. The median follow-up was 6.8 years (range, 3.0-10.3). The 5-year event-free survival was 82±2% [12 toxic deaths, 5 secondary malignancies, 43 non-response/relapse (central nervous system n=9; all received prednisone during induction)]. The event-free survival rate was 80±5% for patients with precursor B-cell lymphoblastic lymphoma, 82±3% for those with T-cell lymphoblastic lymphoma, and 100% for patients with a mixed phenotype. During induction, significantly more grade III/IV toxicities were observed in patients receiving dexamethasone, resulting in significant treatment delays. The number of toxic deaths did not differ significantly. The only variable associated with outcome was performance status at diagnosis. The 90% event-free survival rate for patients with T-cell lymphoblastic lymphoma shown in study NHL-BFM90 was not replicated, mainly due to more toxic deaths and central nervous system relapses. Dexamethasone in induction may prevent central nervous system relapse more effectively than prednisone but produces a higher burden of toxicity. (#NCT00275106).

[Flow-cytometric immunophenotyping in clinical diagnostics]. / Durchflusszytometrische Immunphänotypisierung in der klinischen Diagnostik.

Flow cytometric immunophenotyping represents an indispensable tool in hematological and immunological diagnostics. The most frequent indications include lymphocyte phenotyping and the diagnosis and monitoring of benign and malignant hematologic diseases. The role of immunophenotyping in clinical practice is evolving rapidly. This review provides an overview of its current applications and limitations.

Dexamethasone vs prednisone in induction treatment of pediatric ALL: results of the randomized trial AIEOP-BFM ALL 2000.

Induction therapy for childhood acute lymphoblastic leukemia (ALL) traditionally includes prednisone; yet, dexamethasone may have higher antileukemic potency, leading to fewer relapses and improved survival. After a 7-day prednisone prephase, 3720 patients enrolled on trial Associazione Italiana di Ematologia e Oncologia Pediatrica and Berlin-Frankfurt-Münster (AIEOP-BFM) ALL 2000 were randomly selected to receive either dexamethasone (10 mg/m(2) per day) or prednisone (60 mg/m(2) per day) for 3 weeks plus tapering in induction. The 5-year cumulative incidence of relapse (± standard error) was 10.8 ± 0.7% in the dexamethasone and 15.6 ± 0.8% in the prednisone group (P < .0001), showing the largest effect on extramedullary relapses. The benefit of dexamethasone was partially counterbalanced by a significantly higher induction-related death rate (2.5% vs 0.9%, P = .00013), resulting in 5-year event-free survival rates of 83.9 ± 0.9% for dexamethasone and 80.8 ± 0.9% for prednisone (P = .024). No difference was seen in 5-year overall survival (OS) in the total cohort (dexamethasone, 90.3 ± 0.7%; prednisone, 90.5 ± 0.7%). Retrospective analyses of predefined subgroups revealed a significant survival benefit from dexamethasone only for patients with T-cell ALL and good response to the prednisone prephase (prednisone good-response [PGR]) (dexamethasone, 91.4 ± 2.4%; prednisone, 82.6 ± 3.2%; P = .036). In patients with precursor B-cell ALL and PGR, survival after relapse was found to be significantly worse if patients were previously assigned to the dexamethasone arm. We conclude that, for patients with PGR in the large subgroup of precursor B-cell ALL, dexamethasone especially reduced the incidence of better salvageable relapses, resulting in inferior survival after relapse. This explains the lack of benefit from dexamethasone in overall survival that we observed in the total cohort except in the subset of T-cell ALL patients with PGR. This trial was registered at www.clinicaltrials.gov (BFM: NCT00430118, AIEOP: NCT00613457).

Role of Tyrosine Kinase Inhibitors in Indolent and Other Mature B-Cell Neoplasms.

Targeting tyrosine kinases represents a highly specific treatment approach for different malignancies. This also includes non-Hodgkin lymphoma since it is well known that these enzymes are frequently involved in the lymphomagenesis. Hereby, tyrosine kinases might either be dysregulated intrinsically or be activated within signal transduction pathways leading to tumor survival and growth. Among others, Bruton's tyrosine kinase (Btk) is of particular interest as a potential therapeutic target. Btk is stimulated by B-cell receptor signaling and activates different transcription factors such as nuclear factor κB. The Btk inhibitor ibrutinib has been approved for the treatment of chronic lymphocytic leukemia and mantle-cell lymphoma recently. Numerous clinical trials evaluating this agent in different combinations (eg, with rituximab or classical chemotherapeutic agents) as a treatment option for aggressive and indolent lymphoma are under way. Here, we summarize the role of tyrosine kinase inhibitors in the treatment of indolent and other non-Hodgkin lymphomas (eg, mantle-cell lymphoma).

Minimal residual disease analysis by eight-color flow cytometry in relapsed childhood acute lymphoblastic leukemia.

Multiparametric flow cytometry is an alternative approach to the polymerase chain reaction method for evaluating minimal residual disease in treatment protocols for primary acute lymphoblastic leukemia. Given considerable differences between primary and relapsed acute lymphoblastic leukemia treatment regimens, flow cytometric assessment of minimal residual disease in relapsed leukemia requires an independent comprehensive investigation. In the present study we addressed evaluation of minimal residual disease by flow cytometry in the clinical trial for childhood relapsed acute lymphoblastic leukemia using eight-color flow cytometry. The major challenge of the study was to reliably identify low amounts of residual leukemic cells against the complex background of regeneration, characteristic of follow-up samples during relapse treatment. In a prospective study of 263 follow-up bone marrow samples from 122 patients with B-cell precursor acute lymphoblastic leukemia, we tested various B-cell markers, adapted the antibody panel to the treatment protocol, and evaluated its performance by a blinded parallel comparison with the polymerase chain reaction data. The resulting eight-color single-tube panel showed a consistently high overall concordance (P<0.001) and, under optimal conditions, sensitivity similar to that of the reference polymerase chain reaction method. Overall, evaluation of minimal residual disease by flow cytometry can be successfully integrated into the clinical management of relapsed childhood acute lymphoblastic leukemia either as complementary to the polymerase chain reaction or as an independent risk stratification tool. ALL-REZ BFM 2002 clinical trial information: NCT00114348.

Flow diagnostics essential code: a simple and brief format for the summary of leukemia phenotyping.

BACKGROUND: Flow cytometry is a valuable part in the routine diagnostics of acute leukemia (AL). Although internationally recognized definitions of main AL subsets are available, there is currently no consensus format for the short summary of clinical flow cytometry reports. Since clinical reports are too long for most database purposes, there is a need for a standardized format of their short summaries. METHODS: The Associazione Italiana Ematologia Oncologia Pediatrica--Berlin Frankfurt Muenster (AIEOP-BFM) Flow Network that encompasses reference diagnostics laboratories in Australia, Austria, Czechia, Germany, Israel, Italy, and Switzerland have designed a pro-forma for the summary of flow cytometry results in the diagnosis of leukemia. The process involved several meetings and other communications, during which the group established a consensus on the essentials that lead to the diagnostic conclusions in childhood AL. RESULTS: The "Flow Diagnostics Essential (FDE) Code" is a result from an agreement within the AIEOP-BFM Flow Network. In a standardized format, it reports the extent of the infiltration by a malignant clone, followed by description antigen expression as strong, weak or negative, and a diagnostic conclusion. CONCLUSIONS: A consensus brief format (the "FDE Code") has been designed as a brief summary of the diagnostic immunophenotype of childhood AL. It is also applicable for the diagnostic investigation of other malignancies by flow cytometry. The FDE code may be included in the final clinical report and/or used in the setting of a multicenter clinical trial database.

High CD45 surface expression determines relapse risk in children with precursor B-cell and T-cell acute lymphoblastic leukemia treated according to the ALL-BFM 2000 protocol.

Further improvement of outcome in childhood acute lymphoblastic leukemia could be achieved by identifying additional high-risk patients who may benefit from intensified treatment. We earlier identified PTPRC (CD45) gene expression as a potential new stratification marker and now analyzed the prognostic relevance of CD45 protein expression. CD45 was measured by flow cytometry in 1065 patients treated according to the ALL-BFM-2000 protocol. The 75(th) percentile was used as cut-off to distinguish a CD45-high from a CD45-low group. As mean CD45 expression was significantly higher in T-cell acute lymphoblastic leukemia than in B-cell-precursor acute lymphoblastic leukemia (P<0.0001), the analysis was performed separately in both groups. In B-cell-precursor acute lymphoblastic leukemia we observed a significant association of a high CD45 expression with older age, high initial white blood cell count, ETV6/RUNX1 negativity, absence of high hyperdiploidy (P<0.0001), MLL/AF4 positivity (P=0.002), BCR/ABL1 positivity (P=0.007), prednisone poor response (P=0.002) and minimal residual disease (P<0.0001). In T-cell acute lymphoblastic leukemia we observed a significant association with initial white blood cell count (P=0.0003), prednisone poor response (P=0.01), and minimal residual disease (P=0.02). Compared to CD45-low patients, CD45-high patients had a lower event-free survival rate (B-cell-precursor acute lymphoblastic leukemia: 72 ± 3% versus 86 ± 1%, P<0.0001; T-cell acute lymphoblastic leukemia: 60 ± 8% versus 78 ± 4%, P=0.02), which was mainly attributable to a higher cumulative relapse incidence (B-cell-precursor acute lymphoblastic leukemia: 22 ± 3% versus 11 ± 1%, P<0.0001; T-cell acute lymphoblastic leukemia: 31 ± 8% versus 11 ± 3%, P=0.003) and kept its significance in multivariate analysis considering sex, age, initial white blood cell count, and minimal residual disease in B-cell-precursor- and T-cell acute lymphoblastic leukemia, and additionally presence of ETV6/RUNX1, MLL/AF4 and BCR/ABL1 rearrangements in B-cell-precursor acute lymphoblastic leukemia (P=0.002 and P=0.025, respectively). Consideration of CD45 expression may serve as an additional stratification tool in BFM-based protocols. (ClinicalTrials.gov identifier: NCT00430118).

Peripheral blood sCD3⁻ CD4⁺ T cells: a useful diagnostic tool in angioimmunoblastic T cell lymphoma.

Angioimmunoblastic T cell lymphoma (AITL) belongs to the subgroup of mature T cell lymphomas according to the World Health Organization and is one of the common T cell lymphomas in Western countries. Particularly in cases in which histological confirmation cannot be easily achieved, immunophenotyping of peripheral blood can give important information for the differential diagnosis of AITL. sCD3⁻ CD4⁺ T cells are a typical feature of AILT in flow cytometry of peripheral blood. In this retrospective study, the diagnostic value of flow cytometry for the diagnosis 'AITL' was assessed by comparing the frequency of sCD3⁻ CD4⁺ T cells in leukemic AITL patients and in patients with other leukemic CD4⁺ T cell lymphomas. Immunophenotyping of peripheral blood by flow cytometry was performed in a lymphocyte gate using fluorochrome-labelled antibodies against CD3, CD2, CD4, CD5, CD7, CD8, CD10, CD14, CD16, CD19, CD56, CD57 and T cell receptor. In 17/17 leukemic AITL patients, a small but distinct population of sCD3⁻ CD4⁺ T cells was detected (mean percentage of sCD3⁻ CD4⁺ T cells in the lymphocyte gate: 11.9 ± 15.4%, range 0.1-51.8%). In contrast, sCD3⁻ CD4⁺ T cells were found in only 1/40 patients with other leukemic CD4⁺ T cell lymphomas (one patient with mycosis fungoides). sCD3⁻ CD4⁺ T cells have a high positive predictive value (94%) for the diagnosis 'AITL'. Flow cytometry is particularly useful in the differential diagnosis of AITL, even if the aberrant T cell population has a very low frequency. Further biological characterization of this subfraction of lymphoma cells is warranted.

IKZF1 deletion is an independent predictor of outcome in pediatric acute lymphoblastic leukemia treated according to the ALL-BFM 2000 protocol.

IKZF1 gene deletions have been associated with a poor outcome in pediatric precursor B-cell acute lymphoblastic leukemia. To assess the prognostic relevance of IKZF1 deletions for patients treated on Berlin-Frankfurt-Münster Study Group trial ALL-BFM 2000, we screened 694 diagnostic acute lymphoblastic leukemia samples by Multiplex Ligation-dependent Probe Amplification. Patients whose leukemic cells bore IKZF1 deletions had a lower 5-year event-free survival (0.69±0.05 vs. 0.85±0.01; P<0.0001) compared to those without, mainly due to a higher cumulative incidence of relapses (0.21±0.04 vs. 0.10±0.01; P=0.001). Although IKZF1 deletions were significantly associated with the P2RY8-CRLF2 rearrangement, their prognostic value was found to be independent from this association. Thus, IKZF1 deletion is an independent predictor of treatment outcome and a strong candidate marker for integration in future treatment stratification strategies on ALL-BFM protocols. Clinicaltrials.gov identifier: NCT00430118.