The Gene Expression Classifier ALLCatchR Identifies B-cell Precursor ALL Subtypes and Underlying Developmental Trajectories Across Age

Deciphering IGH rearrangement complexity and detection strategies in acute lymphoblastic leukaemia

Acute lymphoblastic leukaemia is a highly heterogeneous malignancy characterised by various genomic alterations that influence disease progression and therapeutic outcomes. Gene fusions involving the immunoglobulin heavy chain gene represent a complex and diverse category. These fusions often result in enhancer hijacking, upregulation of partner proto-oncogenes and contribute to leukemogenesis. This review highlights the mechanisms underlying IGH gene fusions, the critical role they play in ALL pathogenesis, and current detection technologies.

Genetic subtypes of B-cell acute lymphoblastic leukemia in adults

ABSTRACT

B-cell acute lymphoblastic leukemia (B-ALL) is a rare malignancy in adults, with outcomes remaining poor, especially compared with children. Over the past 2 decades, extensive whole-genome studies have identified numerous genetic alterations driving leukemia, leading to the recognition of >20 distinct subtypes that are closely associated with treatment response and prognosis. In pediatric B-ALL, large correlation studies have made genetic classification a central component of risk-adapted treatment strategies. Notably, genetic subtypes are unevenly distributed according to age, and the spectrum of genetic alterations and their prognostic relevance in adult B-ALL have been less extensively studied, with treatment primarily based on the presence or absence of BCR::ABL1 fusion. This review provides an overview of genetic subtypes in adult B-ALL, including recent biological and clinical insights in well-established subtypes as well as data on newly recognized subtypes. Their relevance for risk classification, disease monitoring, and therapeutic management, including in the context of B-cell-directed therapies, is discussed. This review advocates for continuing efforts to further improve our understanding of the biology of adult B-ALL to establish the foundation of future precision medicine in B-ALL.

RNA-sequencing: A reliable tool to unveil transcriptional landscape of paediatric B-other acute lymphoblastic leukaemia

B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) comprises multiple subtypes characterized by different genetic alterations. With the use of current standard-of-care tests used in clinical practice, 20%-30% of the cases may not be classified into the main genetic subtypes and additional approaches are needed. These patients are grouped in the heterogeneous category B-other ALL. Transcriptome sequencing (RNA-seq) has allowed the identification of novel fusion genes and gene expression profiles that define new molecular subtypes. We present RNA-seq results integrated, in a real-world scenario, with clinical routine diagnostic data to identify new biomarkers and reclassify a cohort of 60 B-other ALL patients in the newly described genetic subtypes. Overall, 49 rearrangements were identified, including 32 different fusion genes in 41 B-other patients (68%). Moreover, we reported six novel rearrangements (IGK::PAX5, PAX5::IL1RAPL1, ETV6::KRT78, IGH::HIC1, IGH::MIR100HG and NKAIN4::PNPLA7). The integration of RNA-seq results with standard-of-care data allowed us to classify 72% of the patients (43/60) in 11 different subtypes, being DUX4 rearranged and PAX5alt the most represented subtypes. In summary, RNA-seq is a reliable tool for the identification of new emerging genetic subtypes contributing to a better genetic risk stratification of BCP-ALL paediatric patients on the path towards a more personalized medicine.

Application of Omics Analyses in Pediatric B-Cell Acute Lymphoblastic Leukemia

Acute lymphoblastic leukemia (ALL) is the most common pediatric cancer, comprising almost 25% of all malignancies diagnosed in children younger than 20 years, and its incidence is still increasing. ALL is a blood cancer arising from the unregulated proliferation of clonal lymphoid progenitor cells. To make a diagnosis of B-cell ALL, bone marrow morphology and immunophenotyping are needed; cerebrospinal fluid examination, and chromosomal analysis are currently used as stratification exams. Currently, almost 70% of children affected by B-cell ALL are characterized by well-known cytogenetic abnormalities. However, the integration of results with "omic" techniques (genomics, transcriptomics, proteomics, and metabolomics, both individually and integrated) able to analyze simultaneously thousands of molecules, has enabled a deeper definition of the molecular scenario of B-cell ALL and the identification of new genetic alterations. Studies based on omics have greatly deepened our knowledge of ALL, expanding the horizon from the traditional morphologic and cytogenetic point of view. In this review, we focus our attention on the "omic" approaches mainly used to improve the understanding and management of B-cell ALL, crucial for the diagnosis, prognosis, and treatment of the disease, offering a pathway toward more precise and personalized therapeutic interventions.

Challenging Conventional Diagnostic Methods by Comprehensive Molecular Diagnostics: A Nationwide Prospective Comparison in Children With ALL

PURPOSE

Treatment stratification in ALL includes diverse (cyto)genetic aberrations, requiring diverse tests to yield conclusive data. We optimized the diagnostic workflow to detect all relevant aberrations with a limited number of tests in a clinically relevant time frame.

METHODS

In 467 consecutive patients with ALL (0-20 years), we compared RNA sequencing (RNAseq), fluorescence in situ hybridization (FISH), reverse transcriptase polymerase chain reaction (RT-PCR), karyotyping, single-nucleotide polymorphism (SNP) array, and multiplex ligation-dependent probe amplification (MLPA) for technical success, concordance of results, and turnaround time.

RESULTS

To detect stratifying fusions (ETV6::RUNX1, BCR::ABL1, ABL-class, KMT2Ar, TCF3::HLF, IGH::MYC), RNAseq and FISH were conclusive for 97% and 96% of patients, respectively, with 99% concordance. RNAseq performed well in samples with a low leukemic cell percentage or low RNA quality. RT-PCR for six specific fusions was conclusive for >99% but false-negative for six patients with alternatively fused exons. RNAseq also detected gene fusions not yet used for stratification in 14% of B-cell precursor-ALL and 33% of T-ALL. For aneuploidies and intrachromosomal amplification of chromosome 21, SNP array gave a conclusive result in 99%, thereby outperforming karyotyping, which was conclusive for 64%. To identify deletions in eight stratifying genes/regions, SNP array was conclusive in 99% and MLPA in 95% of patients, with 98% concordance. The median turnaround times were 10 days for RNAseq, 9 days for FISH, 10 days for SNP array, and <7 days for MLPA and RT-PCR in this real-world prospective study.

CONCLUSION

Combining RNAseq and SNP array outperformed current diagnostic tools to detect all stratifying genetic aberrations in ALL. The turnaround time is <15 days matching major treatment decision time points. Moreover, combining RNAseq and SNP array has the advantage of detecting new lesions for studies on prognosis and pathobiology.

RanBALL: An Ensemble Random Projection Model for Identifying Subtypes of B-Cell Acute Lymphoblastic Leukemia

As the most common pediatric malignancy, B-cell acute lymphoblastic leukemia (B-ALL) has multiple distinct subtypes characterized by recurrent and sporadic somatic and germline genetic alterations. Identifying B-ALL subtypes can facilitate risk stratification and enable tailored therapeutic design. Existing methods for B-ALL subtyping primarily depend on immunophenotyping, cytogenetic tests and genomic profiling, which would be costly, complicated, and laborious. To overcome these challenges, we present RanBALL (an ensemble Random projection-based model for identifying B-ALL subtypes), an accurate and cost-effective model for B-ALL subtype identification. By leveraging random projection (RP) and ensemble learning, RanBALL can preserve patient-to-patient distances after dimension reduction and yield robustly accurate classification performance for B-ALL subtyping. Benchmarking results based on > 1700 B-ALL patients demonstrated that RanBALL achieved remarkable performance (accuracy: 0.93, F1-score: 0.93, and Matthews correlation coefficient: 0.93), significantly outperforming state-of-the-art methods like ALLSorts in terms of all performance metrics. In addition, RanBALL performs better than tSNE in terms of visualizing B-ALL subtype information. We believe RanBALL will facilitate the discovery of B-ALL subtype-specific marker genes and therapeutic targets to have consequential positive impacts on downstream risk stratification and tailored treatment design. To extend its applicability and impacts, a Python-based RanBALL package is available at https://github.com/wan-mlab/RanBALL.

SOHO State of the Art Updates and Next Questions | Approach to BCR::ABL1-Like Acute Lymphoblastic Leukemia

Philadelphia-like (Ph-like) or BCR::ABL1-like acute lymphoblastic leukemia (ALL) is a common high-risk subtype of B-cell precursor ALL (B-ALL) characterized by a diverse range of genetic alterations that challenge diagnose and converge on distinct kinase and cytokine receptor-activated gene expression profiles, resembling those from BCR::ABL1-positive ALL from which its nomenclature. The presence of kinase-activating genetic drivers has prompted the investigation in preclinical models and clinical settings of the efficacy of tyrosine kinase inhibitor (TKI)-based treatments. This was further supported by an inadequate response to conventional chemotherapy, high rates of induction failure and persistent measurable residual disease (MRD) positivity, which translate in lower survival rates compared to other B-ALL subtypes. Therefore, innovative approaches are underway, including the integration of TKIs with frontline regimens and the early introduction of immunotherapy strategies (monoclonal antibodies, T-cell engagers, drug-conjugates, and CAR-T cells). Allogeneic hematopoietic cell transplantation (HSCT) is currently recommended for adult BCR::ABL1-like ALL patients in first complete remission. However, the incorporation of novel therapies, a more accurate diagnosis and a more sensitive MRD assessment may modify the risk stratification and the indication for transplant in these patients.

Microfluidic Affinity Selection of B-Lineage Cells from Peripheral Blood for Minimal Residual Disease Monitoring in Pediatric B-Type Acute Lymphoblastic Leukemia Patients

Assessment of minimal residual disease (MRD) is the most powerful predictor of outcome in B-type acute lymphoblastic leukemia (B-ALL). MRD, defined as the presence of leukemic cells in the blood or bone marrow, is used for the evaluation of therapy efficacy. We report on a microfluidic-based MRD (MF-MRD) assay that allows for frequent evaluation of blood for the presence of circulating leukemia cells (CLCs). The microfluidic chip affinity selects B-lineage cells, including CLCs using anti-CD19 antibodies poised on the wall of the microfluidic chip. Affinity-selected cells are released from the capture surface and can be subjected to immunophenotyping to enumerate the CLCs, perform fluorescence in situ hybridization (FISH), and/or molecular analysis of the CLCs' mRNA/gDNA. During longitudinal testing of 20 patients throughout induction and consolidation therapy, the MF-MRD performed 116 tests, while only 41 were completed with multiparameter flow cytometry (MFC-MRD) using a bone marrow aspirate, as standard-of-care. Overall, 57% MF-MRD tests were MRD(+) as defined by CLC numbers exceeding a threshold of 5 × 10-4%, which was determined to be the limit of quantitation. Above a threshold of 0.01%, MFC-MRD was positive in 34% of patients. The MF offered the advantage of the opportunity for efficiently processing small volumes of blood (2 mL), which is important in the care of pediatric patients, especially infants. The minimally invasive means of blood collection are of high value when treating patients whose MRD is typically tested using an invasive bone marrow biopsy. MF-MRD detection can be useful for stratification of patients into risk groups and monitoring of patient well-being after completion of treatment for early recognition of potential impending disease recurrence.

Mutational and transcriptional landscape of pediatric B-cell precursor lymphoblastic lymphoma

ABSTRACT

Pediatric B-cell precursor (BCP) lymphoblastic malignancies are neoplasms with manifestation either in the bone marrow or blood (BCP acute lymphoblastic leukemia [BCP-ALL]) or are less common in extramedullary tissue (BCP lymphoblastic lymphoma [BCP-LBL]). Although both presentations are similar in morphology and immunophenotype, molecular studies have been virtually restricted to BCP-ALL so far. The lack of molecular studies on BCP-LBL is due to its rarity and restriction on small, mostly formalin-fixed paraffin-embedded (FFPE) tissues. Here, to our knowledge, we present the first comprehensive mutational and transcriptional analysis of what we consider the largest BCP-LBL cohort described to date (n = 97). Whole-exome sequencing indicated a mutational spectrum of BCP-LBL, strikingly similar to that found in BCP-ALL. However, epigenetic modifiers were more frequently mutated in BCP-LBL, whereas BCP-ALL was more frequently affected by mutation in genes involved in B-cell development. Integrating copy number alterations, somatic mutations, and gene expression by RNA sequencing revealed that virtually all molecular subtypes originally defined in BCP-ALL are present in BCP-LBL, with only 7% of lymphomas that were not assigned to a subtype. Similar to BCP-ALL, the most frequent subtypes of BCP-LBL were high hyperdiploidy and ETV6::RUNX1. Tyrosine kinase/cytokine receptor rearrangements were detected in 7% of BCP-LBL. These results indicate that genetic subtypes can be identified in BCP-LBL using next-generation sequencing, even in FFPE tissue, and may be relevant to guide treatment.

Acute lymphoblastic leukaemia

Acute lymphoblastic leukaemia (ALL) is a haematological malignancy characterized by the uncontrolled proliferation of immature lymphoid cells. Over past decades, significant progress has been made in understanding the biology of ALL, resulting in remarkable improvements in its diagnosis, treatment and monitoring. Since the advent of chemotherapy, ALL has been the platform to test for innovative approaches applicable to cancer in general. For example, the advent of omics medicine has led to a deeper understanding of the molecular and genetic features that underpin ALL. Innovations in genomic profiling techniques have identified specific genetic alterations and mutations that drive ALL, inspiring new therapies. Targeted agents, such as tyrosine kinase inhibitors and immunotherapies, have shown promising results in subgroups of patients while minimizing adverse effects. Furthermore, the development of chimeric antigen receptor T cell therapy represents a breakthrough in ALL treatment, resulting in remarkable responses and potential long-term remissions. Advances are not limited to treatment modalities alone. Measurable residual disease monitoring and ex vivo drug response profiling screening have provided earlier detection of disease relapse and identification of exceptional responders, enabling clinicians to adjust treatment strategies for individual patients. Decades of supportive and prophylactic care have improved the management of treatment-related complications, enhancing the quality of life for patients with ALL.

An artificial intelligence-assisted clinical framework to facilitate diagnostics and translational discovery in hematologic neoplasia

BACKGROUND

The increasing volume and intricacy of sequencing data, along with other clinical and diagnostic data, like drug responses and measurable residual disease, creates challenges for efficient clinical comprehension and interpretation. Using paediatric B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) as a use case, we present an artificial intelligence (AI)-assisted clinical framework clinALL that integrates genomic and clinical data into a user-friendly interface to support routine diagnostics and reveal translational insights for hematologic neoplasia.

METHODS

We performed targeted RNA sequencing in 1365 cases with haematological neoplasms, primarily paediatric B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) from the AIEOP-BFM ALL study. We carried out fluorescence in situ hybridization (FISH), karyotyping and arrayCGH as part of the routine diagnostics. The analysis results of these assays as well as additional clinical information were integrated into an interactive web interface using Bokeh, where the main graph is based on Uniform Manifold Approximation and Projection (UMAP) analysis of the gene expression data. At the backend of the clinALL, we built both shallow machine learning models and a deep neural network using Scikit-learn and PyTorch respectively.

FINDINGS

By applying clinALL, 78% of undetermined patients under the current diagnostic protocol were stratified, and ambiguous cases were investigated. Translational insights were discovered, including IKZF1plus status dependent subpopulations of BCR::ABL1 positive patients, and a subpopulation within ETV6::RUNX1 positive patients that has a high relapse frequency. Our best machine learning models, LDA and PASNET-like neural network models, achieve F1 scores above 97% in predicting patients' subgroups.

INTERPRETATION

An AI-assisted clinical framework that integrates both genomic and clinical data can take full advantage of the available data, improve point-of-care decision-making and reveal clinically relevant insights promptly. Such a lightweight and easily transferable framework works for both whole transcriptome data as well as the cost-effective targeted RNA-seq, enabling efficient and equitable delivery of personalized medicine in small clinics in developing countries.

FUNDING

German Ministry of Education and Research (BMBF), German Research Foundation (DFG) and Foundation for Polish Science.

MD-ALL: an integrative platform for molecular diagnosis of B-acute lymphoblastic leukemia

B-acute lymphoblastic leukemia (B-ALL) consists of dozens of subtypes defined by distinct gene expression profiles (GEP) and various genetic lesions. With the application of transcriptome sequencing (RNA sequencing [RNA-seq]), multiple novel subtypes have been identified, which lead to an advanced B-ALL classification and risk-stratification system. However, the complexity of analyzing RNA-seq data for B-ALL classification hinders the implementation of the new B-ALL taxonomy. Here, we introduce Molecular Diagnosis of Acute Lymphoblastic Leukemia (MD-ALL), an integrative platform featuring sensitive and accurate B-ALL classification based on GEP and sentinel genetic alterations from RNA-seq data. In this study, we systematically analyzed 2,955 B-ALL RNA-seq samples and generated a reference dataset representing all the reported B-ALL subtypes. Using multiple machine learning algorithms, we identified the feature genes and then established highly sensitive and accurate models for B-ALL classification using either bulk or single-cell RNA-seq data. Importantly, this platform integrates multiple aspects of key genetic lesions acquired from RNA-seq data, which include sequence mutations, large-scale copy number variations, and gene rearrangements, to perform comprehensive and definitive B-ALL classification. Through validation in a hold-out cohort of 974 samples, our models demonstrated superior performance for B-ALL classification compared with alternative tools. Moreover, to ensure accessibility and user-friendly navigation even for users with limited or no programming background, we developed an interactive graphical user interface for this MD-ALL platform, using the R Shiny package. In summary, MD-ALL is a user-friendly B-ALL classification platform designed to enable integrative, accurate, and comprehensive B-ALL subtype classification. MD-ALL is available from https://github.com/gu-lab20/MD-ALL.

Developmental trajectories and cooperating genomic events define molecular subtypes of BCR::ABL1-positive ALL

ABSTRACT

Distinct diagnostic entities within BCR::ABL1-positive acute lymphoblastic leukemia (ALL) are currently defined by the International Consensus Classification of myeloid neoplasms and acute leukemias (ICC): "lymphoid only", with BCR::ABL1 observed exclusively in lymphatic precursors, vs "multilineage", where BCR::ABL1 is also present in other hematopoietic lineages. Here, we analyzed transcriptomes of 327 BCR::ABL1-positive patients with ALL (age, 2-84 years; median, 46 years) and identified 2 main gene expression clusters reproducible across 4 independent patient cohorts. Fluorescence in situ hybridization analysis of fluorescence-activated cell-sorted hematopoietic compartments showed distinct BCR::ABL1 involvement in myeloid cells for these clusters (n = 18/18 vs n = 3/16 patients; P < .001), indicating that a multilineage or lymphoid BCR::ABL1 subtype can be inferred from gene expression. Further subclusters grouped samples according to cooperating genomic events (multilineage: HBS1L deletion or monosomy 7; lymphoid: IKZF1-/- or CDKN2A/PAX5 deletions/hyperdiploidy). A novel HSB1L transcript was highly specific for BCR::ABL1 multilineage cases independent of HBS1L genomic aberrations. Treatment on current German Multicenter Study Group for Adult ALL (GMALL) protocols resulted in comparable disease-free survival (DFS) for multilineage vs lymphoid cluster patients (3-year DFS: 70% vs 61%; P = .530; n = 91). However, the IKZF1-/- enriched lymphoid subcluster was associated with inferior DFS, whereas hyperdiploid cases showed a superior outcome. Thus, gene expression clusters define underlying developmental trajectories and distinct patterns of cooperating events in BCR::ABL1-positive ALL with prognostic relevance.

Transcriptome Sequencing Allows Comprehensive Genomic Characterization of Pediatric B-Acute Lymphoblastic Leukemia in an Academic Clinical Laboratory

Studies have shown the power of transcriptome sequencing [RNA sequencing (RNA-Seq)] in identifying known and novel oncogenic drivers and molecular subtypes of B-acute lymphoblastic leukemia (B-ALL). The current study investigated whether the clinically validated RNA-Seq assay, coupled with a custom analysis pipeline, could be used for a comprehensive B-ALL classification. Following comprehensive clinical testing, RNA-Seq was performed on 76 retrospective B-ALL cases, 28 of which had known and 48 had undetermined subtype. Subtypes were accurately identified in all 28 known cases, and in 38 of 48 unknown cases (79%). The subtypes of the unknown cases included the following: PAX5alt (n = 12), DUX4-rearranged (n = 6), Philadelphia chromosome-like (n = 5), low hyperdiploid (n = 4), ETV6::RUNX1-like (n = 3), MEF2D-rearranged (n = 2), PAX5 P80R (n = 2), ZEB2/CEBP (n = 1), NUTM1-rearranged (n = 1), ZNF384-rearranged (n = 1), and TCF3::PBX1 (n = 1). In 15 of 38 cases (39%), classification based on expression profile was corroborated by detection of subtype-defining oncogenic drivers missed by clinical testing. RNA-Seq analysis also detected large copy number abnormalities, oncogenic hot-spot sequence variants, and intragenic IKZF1 deletions. This pilot study confirms the feasibility of implementing an RNA-Seq workflow for clinical diagnosis of molecular subtypes in pediatric B-ALL, reinforcing that RNA-Seq represents a promising global genomic assay for this heterogeneous leukemia.

Multimodal classification of molecular subtypes in pediatric acute lymphoblastic leukemia

Genomic analyses have redefined the molecular subgrouping of pediatric acute lymphoblastic leukemia (ALL). Molecular subgroups guide risk-stratification and targeted therapies, but outcomes of recently identified subtypes are often unclear, owing to limited cases with comprehensive profiling and cross-protocol studies. We developed a machine learning tool (ALLIUM) for the molecular subclassification of ALL in retrospective cohorts as well as for up-front diagnostics. ALLIUM uses DNA methylation and gene expression data from 1131 Nordic ALL patients to predict 17 ALL subtypes with high accuracy. ALLIUM was used to revise and verify the molecular subtype of 281 B-cell precursor ALL (BCP-ALL) cases with previously undefined molecular phenotype, resulting in a single revised subtype for 81.5% of these cases. Our study shows the power of combining DNA methylation and gene expression data for resolving ALL subtypes and provides a comprehensive population-based retrospective cohort study of molecular subtype frequencies in the Nordic countries.